Università degli Studi di Padova

DIPARTIMENTO DI INGEGNERIA INDUSTRIALE DII

Corso di Laurea Magistrale in Ingegneria dell’Energia Elettrica

Tesi di laurea magistrale

Advanced LiDAR Systems

Design of a LiDAR platform

Candidato: Relatore:
Vittorio Giarola Prof. Arturo Lorenzoni
Matricola 1152954
Correlatore:
Prof. Wei Wei

Anno Accademico 2017–2018

 iii

To my Family
iv
 Abstract

It’s possible to say that one of the most interesting laser application is Li-
DAR. This system can measure the distance between a laser source and a
target by hitting it with a pulsed light and then revealing the reflected ray
with a suitable detector. Differences in laser return times and wavelengths
can then be used to make digital 3D-representations of the target. This is
the basic idea behind this complex and fascinating object. The willing of
understanding better its way of functioning and exploiting modern applica-
tions in obstacles detection made me curious and at the same time excited,
to work in this research field. The final aim of this work is to show that is
possible to design and manufacture a LiDAR prototype that associate good
performances and a low price; typically less than 250 dollars. Moreover, this
work promises to develop an ecosystem able to show the LiDAR data in real
time. So, in other words, an integrated environment.
vi
Contents

1 Project Development 1
1.1 Why Project Management . . . . . . . . . . . . . . . . . . . . 1
1.2 The procedure . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Introduction on LiDAR Systems 7

2.1 What is a LiDAR . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Lidar Applications . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.1 Airborn LiDAR . . . . . . . . . . . . . . . . . . . . . . 11
2.2.2 Terrestrial LiDAR . . . . . . . . . . . . . . . . . . . . 12
2.2.3 Autonomous vehicles . . . . . . . . . . . . . . . . . . . 13
2.2.4 Urban Planning . . . . . . . . . . . . . . . . . . . . . . 14
2.2.5 Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3 State of the art LiDAR scanning Technology . . . . . . . . . 15
2.4 LiDAR Project Scope . . . . . . . . . . . . . . . . . . . . . . 19

3 Fundamentals on LiDAR 21
3.1 The Sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1.1 Technology and Operations . . . . . . . . . . . . . . . 22
3.1.2 Interface . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 Arduino . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2.1 Programming . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.2 Connection Between Arduino and the Sensor . . . . . 29
3.2.3 Power . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2.4 Input and Outputs . . . . . . . . . . . . . . . . . . . . 32
3.3 Servo Motors . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.3.1 How It Works . . . . . . . . . . . . . . . . . . . . . . . 34
3.3.2 Control Strategy . . . . . . . . . . . . . . . . . . . . . 36
3.4 Processing v3 . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.4.1 Export . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4 LiDAR Platform v1 41
4.1 Sensor control Strategy . . . . . . . . . . . . . . . . . . . . . 43
4.2 Servo control Strategy . . . . . . . . . . . . . . . . . . . . . . 46

vii
viii CONTENTS

4.2.1 Servo characteristic determination . . . . . . . . . . . 47

4.3 Data Elaboration and Transmission . . . . . . . . . . . . . . . 48
4.4 Data Processing, Visualization and Results . . . . . . . . . . 53
4.5 Towards LiDAR v2 . . . . . . . . . . . . . . . . . . . . . . . . 58

5 LiDAR platform v2 63
5.1 Stepper Motors . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.1.1 Micro Stepper driver and wiring scheme . . . . . . . . 68
5.2 Experiment Set up . . . . . . . . . . . . . . . . . . . . . . . . 70
5.2.1 Camera specifications and features . . . . . . . . . . . 70
5.2.2 Carbon fiber single layer pre-preg . . . . . . . . . . . . 71
5.2.3 Equipment List . . . . . . . . . . . . . . . . . . . . . . 71
5.3 Experimental Data . . . . . . . . . . . . . . . . . . . . . . . . 75
5.3.1 Data Report . . . . . . . . . . . . . . . . . . . . . . . 75
5.3.2 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 80
5.4 New Structure . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.5 Results and Scan . . . . . . . . . . . . . . . . . . . . . . . . . 81

6 Conclusions 87
List of Figures

1.1 Adaptive PM process . . . . . . . . . . . . . . . . . . . . . . . 2

1.2 Table of requirements . . . . . . . . . . . . . . . . . . . . . . 3
1.3 The process . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Timeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.1 LiDAR, working principle . . . . . . . . . . . . . . . . . . . . 8

2.2 3D flash LIDAR by Continental AG [13] . . . . . . . . . . . . 16
2.3 Leica BLK360 . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

3.1 LiDAR Lite I2C read/write operations . . . . . . . . . . . . . 26

3.2 Arduino IDE . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.3 Aduino-LiDAR Connection . . . . . . . . . . . . . . . . . . . 31
3.4 Arduino ATMega Chipset . . . . . . . . . . . . . . . . . . . . 32
3.5 Servo Motor, scheme . . . . . . . . . . . . . . . . . . . . . . . 34
3.6 Stepper Gear examples . . . . . . . . . . . . . . . . . . . . . . 35
3.7 Servo sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.8 Servo Control Strategy . . . . . . . . . . . . . . . . . . . . . . 37
3.9 Processing v3 IDE . . . . . . . . . . . . . . . . . . . . . . . . 39

4.1 Timeline detail v1 . . . . . . . . . . . . . . . . . . . . . . . . 42

4.2 LiDAR v1 Architecture . . . . . . . . . . . . . . . . . . . . . 43
4.3 Servo lab test 1 . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.4 Servo lab test 2 . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.5 Servo Characteristic, graph . . . . . . . . . . . . . . . . . . . 51
4.6 Spherical Coordinates . . . . . . . . . . . . . . . . . . . . . . 52
4.7 Spherical & Cartesian Coordinates . . . . . . . . . . . . . . . 52
4.8 Scan 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.9 Scan 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.10 Scan 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

5.1 Experiment step by step . . . . . . . . . . . . . . . . . . . . . 64

5.2 Working principle Stepper (a) . . . . . . . . . . . . . . . . . . 66
5.3 Working principle Stepper (b) . . . . . . . . . . . . . . . . . . 66
5.4 Working principle Stepper (micro stepping) . . . . . . . . . . 67

ix
x LIST OF FIGURES

5.5 Bipolar Stepper . . . . . . . . . . . . . . . . . . . . . . . . . . 67

5.6 Micro Stepper Driver (DM320) . . . . . . . . . . . . . . . . . 69
5.7 Stepper Module . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.8 Camera drawing & features table . . . . . . . . . . . . . . . . 71
5.9 Carbon fiber pre-preg sample . . . . . . . . . . . . . . . . . . 72
5.10 Experimental Set up . . . . . . . . . . . . . . . . . . . . . . . 72
5.11 Computational Process . . . . . . . . . . . . . . . . . . . . . . 76
5.12 Measurement Sample . . . . . . . . . . . . . . . . . . . . . . . 77
5.13 ∆/Ns vs Ns (1/2) . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.14 ∆/Ns vs Ns (1/2) - Detail . . . . . . . . . . . . . . . . . . . . . 79
5.15 ∆/Ns vs Ns (1/4) . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.16 Carbon Fiber Sample Detail . . . . . . . . . . . . . . . . . . . 80
5.17 Mechanical Assembly, Detail . . . . . . . . . . . . . . . . . . 82
5.18 LiDAR v2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.19 Table and Boxes Scan, v2 . . . . . . . . . . . . . . . . . . . . 83
5.20 Living Room Scan, v2 . . . . . . . . . . . . . . . . . . . . . . 84
5.21 Human Body Scan . . . . . . . . . . . . . . . . . . . . . . . . 85
List of Tables

2.1 BLK360 Laser Specs . . . . . . . . . . . . . . . . . . . . . . . 17

3.1 Phisical Propriety Garmin Lidar Lite v3 . . . . . . . . . . . . 22

3.2 Electrical Propriety Garmin Lidar Lite v3 . . . . . . . . . . . 22
3.3 Garmin Lidar Lite v3 Performances . . . . . . . . . . . . . . 27
3.4 Laser Propriety Garmin Lidar Lite v3 . . . . . . . . . . . . . 27
3.5 Arduino MCU specs . . . . . . . . . . . . . . . . . . . . . . . 28
3.6 Arduino-LiDAR Connection . . . . . . . . . . . . . . . . . . . 31

4.1 Servo Characteristic . . . . . . . . . . . . . . . . . . . . . . . 49

4.2 Table of enhancement . . . . . . . . . . . . . . . . . . . . . . 59

5.1 Camera Features . . . . . . . . . . . . . . . . . . . . . . . . . 70

5.2 Equipment List . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.3 Data Set 1, fms = 1/2 . . . . . . . . . . . . . . . . . . . . . . . 78
5.4 Data Set 2, fms = 1/4 . . . . . . . . . . . . . . . . . . . . . . . 79

xi
xii LIST OF TABLES
Chapter 1

Project Development

Considering the fact that this project is strongly technology and product
oriented we think that it should be better to follow a mind set based on
project management. Considering that, it’s important to walk on the well
known trace of PM, putting a strong attention to the agile method. This
consideration comes from the need of a flexible and constantly adjustable
development strategy to deliver the project even considering time as a main
driver of the process. These are the projects for which the goal is clearly
defined but the solution is not. They are usually called Iterative and Adap-
tive life cycles. We hope that this short introduction has been sufficiently
clear and able to explain the reason why we would like to proceed on that
way.

1.1 Why Project Management

Considering the fact that we are actually developing a brand new product
and we want to deliver something exploitable on the market, there’s the need
of a structure, a general way of thinking, which will help us to fight against
misfortunes and at the same time manage unexpected events. One of the
keys of the project is time, as said. Time is scarce and we need to optimize
and refine its management. For all these reasons it’s a good thing to set up
a structured project thanks to which we can understand from the mistakes
we commit. And then, for sure, it’s even a good thing to learn something
more about modern tools of project management, thinking about a possible
future work experience.
To be honest, the first thing we have to do is to choose which kind of project
management life cycle suits best our situation. Our attention fell on Adap-
tive models which are able to accommodate a high level of uncertainty and
complexity. In that sense, they fill a void between the traditional Iterative
and Extreme models: keeping in mind that solution discovery is still the
focus of these models. Each iteration in the Adaptive models must address

1
2 CHAPTER 1. PROJECT DEVELOPMENT

not only task completion for newly defined functions and features but also
further solution definition through function and feature discovery. It is the
discovery part of the Adaptive PMLC1 models that sets them apart from
other PMLC models.[1] An Adaptive PMLC model consists of a number
of phases that are repeated in cycles, with a feedback loop after each cy-
cle is completed. Each cycle proceeds based on an incomplete and limited
understanding of the solution. Each cycle learns from the preceding cycles
and plans the next cycle in an attempt to converge on an acceptable so-
lution. Adaptive PMLC model is missing both depth and breadth of the
solution. Figure 1.1 depicts the Adaptive PMLC model for projects that
meet the conditions of an incomplete solution due to missing features and
functions. In the Adaptive PMLC model, as with other Agile approaches,
the degree to which the solution is known might vary over a wide range from
knowing a lot but not all. The less that is known about the solution, the
more risk, uncertainty, and complexity will be present. To remove the un-
certainty associated with these projects, the solution has to be discovered.
That will happen through a continuous change process from cycle to cycle.
That change process is designed to create a convergence to a complete solu-
tion. In the absence of that convergence, Adaptive projects are frequently
cancelled and restarted in some other promising direction.

M onitor N ext N
Scope P lan Launch Close Close
Iteration Iteration &Control Iteration Iteration P roject
Iteration

Figure 1.1: Adaptive PM process

1.2 The procedure

Abraham Lincoln was used to say: ”Give me six hours to chop down a
tree and I will spend the first four sharpening the axe.” Since we like this
mind set, we think it’s important to put a big effort from the beginning of
the project, trying to guess the possible evolution and what to expect from
the process itself. Since the product is brand new, we want to experience
something that probably we will struggle to govern.
Anyway, clearly speaking, defining the target of the project and roughly the
time in which we have to deliver the final product it’s probably the most
important thing. Once we have established these things we can focus on
the product itself with its requirements and performances. Only thanks
1
PMLC states for project management life cycle.
1.2. THE PROCEDURE 3

to clear requirements we would be able to understand if we are on the

right path, if we are doing well and if we are on time for the delivery.
The following Fig. 1.2 is only a preliminary Requirement chart based on
macroscopic and general requirements. In the next sections we will develop
a new requirement chart listing the technical performances and features of
the product. So if we are asked to summarize how to develop the project

Figure 1.2: Table of requirements

pointing out all the steps and describing the macroscopic phases of if, we can
easily reply explaining Fig. 1.3. In this picture are reported the milestones
of the project. By the continuous check and monitoring of the status of the
project, taking into account the successful engagement of the milestones we
are able to understand how we are proceeding. In Fig. 1.4 is showed the
timeline of the whole project. The fist step is related to the acquisition of the
informations. At this stage is fundamental to study the LiDAR technology
by itself and then make a research related to the prior art. Talking about
this last topic, we will firstly focus our attention on the market products
and secondly to the open sources projects. The open source community
offers a huge variety of examples and related projects from which we can
take some inspiration. Moreover, the DIY spirit fits perfectly our idea to
keep low the cost of our platform. In the second stage it is expected to
identify the components which are necessary to manufacture the LiDAR, at
least in its first version. This version will be a learning trial version from
which we expect to gather experience and understand the limits and the
boundaries of the system. Once we have a precise idea of the equipment, it
will be necessary to understand how to organize our work and necessarily
4 CHAPTER 1. PROJECT DEVELOPMENT

Step Step Step Step

1 2 3 4
Start End

W hat is
aLidar Lidar0 s Lidartrial Lidar
physics V ersion1.0 2.0
Common V ersion
Lidar Standard
applications Rotating
State of system
the art Settingsand
Description

Figure 1.3: The process

undertake a training phase that will allow us to become at least confident

in the use of software and hardware the project requires. It’s extremely
important to get familiar with the microcontroller architecture and IDE2 ,
how the motors work and how to control them, how to show the data in real
time and write a java code that shows the results. Finally, it is important to
have a precise idea of the integration process that is necessary to follow in
order to put the parts together to deliver a robust and reliable LiDAR. Only
then, the first version can be designed. In our case, the prototype is quite
simple, just a turret. After writing the code and make the hardware work
smoothly and checked the processing interface it’s mandatory to test the
accuracy and reliability of the equipment. The data analysis will provide a
solid understanding on the validation of the process and the performances.
Moreover, the key factor is to underline the limitations of the prototype,
trying to figure out how to fix these problems and improve functionalities in
the further version. Then the second version should be the final one: fixing
the problems observed in the previous one and reaching the desired quality
and accuracy standards.

2
IDE stands for integrated development environment.
1.2. THE PROCEDURE 5

TIMELINE
MARCH [3月] APRIL [4月] MAY [5月] JUNE [6月]

Info Acquisition 03.20

Prior art: market products
and opensource platform
04.10
Comps identification
Search appropriate components
05.10
Learning Phase
Arduino microcontroller
Servo control
Processing
Integration

Lidar Version 1.0 05.31

Code & hardware
Test and data

Lidar Version 2.0 06.31

Code & hardware

Test & data

TASK Mile
Stone

Delivery
SubTask xx.yy Date

Figure 1.4: Timeline

6 CHAPTER 1. PROJECT DEVELOPMENT
Chapter 2

Introduction on LiDAR
Systems

2.1 What is a LiDAR

From the origins of electronics we have always heard about RADAR. This
devices is able to detect the position of objects which occupy the space
thanks electromagnetic waves. Only in recent years the RADAR concept has
been exploited in the optics field, by achieving an optical RADAR called Li-
DAR. In other words, RADAR systems emit radio waves and measure what
bounces back, LiDAR uses light waves. It’s a powerful data collection sys-
tem that provides 3-D information for an area of interest or a project area.
This device is versatile and suitable for different purposes. Especially in the
last 5 or 10 years the scientific community and a lot of companies have been
focused on this product trying to imagine possible applications in different
fields of technology.
If we want to write on paper a more precise definition we can say that:
LiDAR is a surveying method that measures distance to a target by illumi-
nating it with a pulsed laser light, and measuring the reflected pulses with a
sensor. Differences in laser return times and wavelengths can then be used
to make digital 3D-representations of the target.[2]
To understand whether LiDAR might be appropriate for our agency’s sur-
veying project needs, we will need a thorough understanding of the accuracy
and data requirements for the project at hand. By mapping out these criti-
cal components in advance we will be able to get a handle on the appropri-
ate LiDAR solution, including the hardware, software, and ground control.
Someone could ask us why? Because a thorough understanding of the type
of project we are undertaking will help us determine the ground control of
the project. For example, a corridor mapping project will be conducted
differently from a county-wide collection. Or more, for example, a mobile-
mapping collection is a whole different work from a terrestrial or airborne

7
8 CHAPTER 2. INTRODUCTION ON LIDAR SYSTEMS

collection. Lidar uses ultraviolet, visible, or near infrared light to image

objects. It can target a wide range of materials, including non-metallic
objects, rocks, rain, chemical compounds, aerosols, clouds and even single
molecules.[3] But one thing we have to keep always in mind: LiDAR is pretty
powerful, but it is not the solution to everything. It cannot ”see” through
walls or surfaces. It cannot peer through clouds; it is actually foiled by fog;
and it cannot see if there is a dense smoke. Anyway, a narrow laser-beam
can map physical features with very high resolutions. Wavelengths vary to
suit the target: from about 10 micrometers to the UV (approximately 250
nm).[3] Typically light is reflected via backscattering, as opposed to pure re-
flection one might find with a mirror. Different types of scattering are used
for different LiDAR applications: most commonly Rayleigh scattering.[4]
The basic LiDAR scheme is the following (see Fig. 2.1): This scheme is pre-

T RAN SM IT T ER RECEIV ER

Beam OP T ICAL
DET ECT OR

DAT A
ACQU ISIT ION
Laser
Detector

Figure 2.1: LiDAR, working principle

sented only to give a general idea of the concept behind LiDAR technology.
We can recognise:

• Laser : 600–1000 nm lasers are most common for non-scientific ap-

plications. They are inexpensive, but since they can be focused and
easily absorbed by the eye, the maximum power is limited by the need
to make them eye-safe. Eye-safety is often a requirement for most
applications. A common alternative, 1550 nm lasers, are eye-safe at
much higher power levels since this wavelength is not focused by the
eye. Laser settings include the laser repetition rate (which controls
the data collection speed). Pulse length is generally an attribute of
the laser cavity length, the number of passes required through the gain
material, and pulsing speed. Better target resolution is achieved with
shorter pulses, provided the LiDAR receiver detectors and electronics
have sufficient bandwidth;[4]
2.1. WHAT IS A LIDAR 9

• Scanner and optics: How fast images can be developed is also affected
by the speed at which they are scanned. There are several options to
scan the azimuth and elevation, including dual oscillating plane mir-
rors, a combination with a polygon mirror and a dual axis scanner.
Optic choices affect the angular resolution and range that can be de-
tected. A hole mirror or a beam splitter are options to collect a return
signal;

• Photo-detector and receiver electronics: Two main photo-detector tech-

nologies are used in Lidar. The first is solid state photo-detectors, such
as silicon avalanche photo-diodes, or photo-multipliers. The sensitiv-
ity of the receiver is another parameter that has to be balanced in a
Lidar system design;

• Position and navigation systems: LiDAR sensors that are mounted on

mobile platforms such as airplanes or satellites require instrumentation
to determine the absolute position and orientation of the sensor. Such
devices generally include a Global Positioning System receiver and an
Inertial Measurement Unit (IMU).

Talking about traditional and conventional LiDAR is good to get familiar

with some specific language which can help to understand better what just
said:

• Repetition Rate: this is the rate at which the laser is pulsing, and it is
measured in kilohertz (KHz). Laser emits extremely quick pulses, so
it’s important to check that all the equipments, such as trans-receiver
and receiver are able to operate at the maximum laser pulsing speed;

• Scan frequency: while the laser is pulsing, the scanner is oscillating,

or moving back and forth. The scan frequency will tell how fast the
scanner is oscillating. A mobile system has a scanner that rotates
continuously in a 360 degree fashion, but most airborne scanners move
back and forth.

• Scan angle: this is measured in degrees and is the distance that the
scanner moves from one end to the other. We will adjust the angle
depending on the application and the accuracy of the desired data
product;

• Flying attitude: it is not a surprise that the farther the platform is

from the target, the lower the accuracy of the data and the less dense
the points will be that define the target area. That’s why for airborne
systems, the flying attitude is so important. This aspect in our case
it’s less relevant since we are interested in short range applications;
10 CHAPTER 2. INTRODUCTION ON LIDAR SYSTEMS

• Nominal point spacing (NPS): The rule is simple enough the more
points that are hit in the collection, the better we will define the tar-
gets. The point sample spacing varies depending on the application.
It’s also good to keep in mind that LiDAR systems are random sam-
pling systems. Although it’s not possible to determine exactly where
the points are going to hit on the target area. The only thing that is
possible to do is that we can decide how many times the target areas
are going to be hit, so for example we can choose a higher frequency
of points to better define the targets.
• Cross track resolution: this is the spacing of the pulses from the LiDAR
system in the scanning direction, or perpendicular to the direction that
the platform is moving;
• Swath: this is the actual distance of the area of coverage for the LiDAR
system. It can vary depending on the scan angle. Mobile LiDAR has a
swath, too, but it is usually fixed and depends on the particular sensor.
For these systems, though, you might not hear the word ”swath” it
may instead be referred to as the ”area of coverage” and will vary
depending on the repetition rate of the sensor;
• Overlap: it’s the amount of redundant area that is covered between
flight lines or swaths within an area of interest. Overlap is not a wasted
effort, because sometimes it provides more accuracy.[3]
So, the accuracy of the system depends not only on laser’s feature but
even on the proprieties of the revelation and the system in general. Here we
have to be honest and add some more informations. Actually, during the
development of this work, we are basically interested in mechanical scan-
ning systems. Even though mechanical rotational detection systems have
some limits and disadvantages, they’re quite easy to implement and de-
velop. One of the major drawbacks is the fact that the scanning speed is
a limit to the performances of the LiDAR and at the same time reduces
the rotating components life, causing fatalities and premature failure. As
with just about any technology that requires precision you cannot count on
LiDAR collection without proper sensor calibration. All sensors should be
calibrated routinely, and calibration is recommended after every installation
into a platform. How you calibrate depends on the system and the LiDAR
provider, but the results are similar. Terrestrial LiDAR scanners, however,
have self calibrations that take place before every collection. In addition,
all LiDAR systems (whether in the air or on the ground) are initially cal-
ibrated by the manufacturer in a lab and during field trials. Whenever a
hardware component is repaired or replaced on a LiDAR sensor head, the
system should be recalibrated. It is also important to remember that every
further modification and adjustment of the system brings to a revision and
check up status of the new configuration.
2.2. LIDAR APPLICATIONS 11

2.2 Lidar Applications

LiDAR originated in the early 1960s, shortly after the invention of the laser,
and combined laser-focused imaging with the ability to calculate distances by
measuring the time for a signal to return using appropriate sensors and data
acquisition electronics. Its first applications came in meteorology, where the
National Center for Atmospheric Research used it to measure clouds. The
general public became aware of the accuracy and usefulness of LiDAR sys-
tems in 1971 during the Apollo 15 mission, when astronauts used a laser
altimeter to map the surface of the moon. At the very beginning LiDAR
system were predominantly used in high-resolution maps, with applications
in geodesy, geomatics, archaeology, geography, geology, geomorphology, seis-
mology, foretry, atmospheric physics, laser guidance, airborne laser swath
mapping (ALSM), and laser altimetry. All those applications are associated
with a high density and high power of the laser beam. Nowadays, con-
versely, we are seeking applications based on low laser beam power and eye
compatible and safety. By considering that it’s possible to exploit Lidar on
short range, making possible 3D object scanning and object detection and
avoiding collision systems. Actually, we are strongly interested in these last
applications and solutions. In the following part we will discuss about the
different general applications and then in the next section we will pay at-
tention only to industrial and commercial applications in obstacle detection
systems.

2.2.1 Airborn LiDAR

While attached to a plane during flight, creates a 3D point cloud model
of the landscape. This is currently the most detailed and accurate method
of creating digital elevation models, replacing photogrammetry. One major
advantage in comparison with photogrammetry is the ability to filter out re-
flections from vegetation from the point cloud model to create a digital sur-
face model which represents ground surfaces such as rivers, paths, cultural
heritage sites, etc., which are concealed by trees. Within the category of air-
borne LiDAR, there is sometimes a distinction made between high-altitude
and low-altitude applications, but the main difference is a reduction in both
accuracy and point density of data acquired at higher altitudes. Airborne
LiDAR can also be used to create bathymetric models in shallow water.[5]
The main constituents of airborne LiDAR include digital elevation models
(DEM) and digital surface models (DSM). The points and ground points are
the vectors of discrete points while DEM and DSM are interpolated raster
grids of discrete points. The process also involves capturing of digital aerial
photographs. In order to interpret deep seated landslides for example, un-
der the cover of vegetation, scarps, tension cracks or tipped trees air borne
LiDAR is used. Air borne LiDAR digital elevation models can see through
12 CHAPTER 2. INTRODUCTION ON LIDAR SYSTEMS

the canopy of forest cover, perform detailed measurements of scarps, erosion

and tilting of electric poles.[6] Airborne lidar data is processed using a tool-
box called Toolbox for Lidar Data Filtering and Forest Studies (TIFFS) for
lidar data filtering and terrain study software. The data is interpolated to
digital terrain models using the software. The laser is directed at the region
to be mapped and each point’s height above the ground is calculated by
subtracting the original z-coordinate from the corresponding digital terrain
model elevation. Based on this height above the ground the non-vegetation
data is obtained which may include objects such as buildings, electric power
lines, flying birds etc. The rest of the points are treated as vegetation and
used for modeling and mapping. Within each of these plots, lidar metrics
are calculated by calculating statistics such as mean, standard deviation,
skewness, percentiles, quadratic mean etc. ;

2.2.2 Terrestrial LiDAR

also terrestrial laser scanning happen on the Earth’s surface and can be
both stationary or mobile. Stationary terrestrial scanning is most common
as a survey method, for example in conventional topography, monitoring,
cultural heritage documentation and forensics.[7] The 3D point clouds ac-
quired from these types of scanners can be matched with digital images taken
of the scanned area from the scanner’s location to create realistic looking
3D models in a relatively short time when compared to other technologies.
Each point in the point cloud is given the colour of the pixel from the image
taken located at the same angle as the laser beam that created the point.
Mobile LiDAR (also mobile laser scanning) is when two or more scanners are
attached to a moving vehicle to collect data along a path. These scanners
are almost always paired with other kinds of equipment, including GNSS
receivers and IMUs. One example application is surveying streets, where
power lines, exact bridge heights, bordering trees, etc. all need to be taken
into account. Instead of collecting each of these measurements individually
in the field with a tachymeter, a 3D model from a point cloud can be cre-
ated where all of the measurements needed can be made, depending on the
quality of the data collected. This eliminates the problem of forgetting to
take a measurement, so long as the model is available, reliable and has an
appropriate level of accuracy. Terrestrial LiDAR mapping involves a process
of occupancy grid map generation. The process involves an array of cells
divided into grids which employs a process to store the height values when
LiDAR data falls into the respective grid cell. A binary map is then created
by applying a particular threshold to the cell values for further processing.
The next step is to process the radial distance and z-coordinates from each
scan to identify which 3D points correspond to each of the specified grid cell
leading to the process of data formation.[8]
2.2. LIDAR APPLICATIONS 13

2.2.3 Autonomous vehicles

From the official NTSB’s (National Transportation Security Board - US)

website we can actually see that an average of 38000 [12] people per year
loose their life on the american roads due to car accidents and fatalities. To
put that in perspective, that’s more than a Boeing 737 falling out of the
sky five times per week in a year. Could we accept something similar in
the aviation? Probably the answer is no. So, this is the main reason why
LiDAR technology has to be exploited in autonomous driving car; but safety
is not the unique reason. The second theme is related to health and money:
taking into account the average commuter in America, which is around 50
minutes per worker per day, you multiply by the number of daily workers,
120 million, you get 6 billion minutes per day lost by the american workers
sat in traffic. If you compare this time to the the average life expectancy,
around 70 years, you get 162 lifetimes waisted in the traffic jam per day. An
other time, this is not social acceptable. And last but not least important,
there’s a reason related to diversity and inclusion: self driving car could help
to improve the mobility of disable people.
Moreover, as per the latest World Health Organization reports, road traffic
accidents account for 1.25 million deaths each year. In fact, according to an
estimate, by the year 2030 road traffic accidents will be the 7th major cause
of death around the world. It is in this context that the UN’s 2030 Agenda
for Sustainable Development becomes even more important. A target has
been set to reduce the number of traffic accident deaths by 50% by 2020. In
light of these statistics, it is clear to see why the use of LiDAR technology
is critical. Unlike other types of sensors or telematics currently available,
LiDAR has improved capabilities when it comes to the detection of objects,
even in cases where there is a complete absence of light. LiDAR’s features,
which include a blind spot monitoring system, adaptive cruise control, along
with a collision avoidance system and pedestrian protection system, are not
only better than the features of other sensors, but are also far more consis-
tent and reliable. Analysts at Technavio predict that the global automotive
LiDAR sensors market will see a CAGR of more than 34% by 2020.
May use LiDAR for obstacle detection and avoidance to navigate safely
through environments, using rotating laser beams. Cost map or point cloud
outputs from the LiDAR sensor provide the necessary data for robot soft-
ware to determine where potential obstacles exist in the environment and
where the robot is in relation to those potential obstacles. Examples of com-
panies that produce LiDAR sensors commonly used in robotics or vehicle
automation are Sick and Hokuyo. Examples of obstacle detection and avoid-
ance products that leverage LiDAR sensors are the Autonomous Solution,
Inc. Forecast 3D Laser System and Velodyne HDL-64E. This is actually the
world in which we want to dive deeply. So in the next section we will look
at the products now available on the market, painting the ”affresco” of the
14 CHAPTER 2. INTRODUCTION ON LIDAR SYSTEMS

so called prior art. The very first generations of automotive adaptive cruise
control systems used only LiDAR sensors and it is expected a more intense
use of these technologies in the next future.

2.2.4 Urban Planning

in this growing phase of urbanization and industrialization there is an emer-
gent need of proper town planning systems. The 3D city models in urban
areas are essential for many applications, such as military operations, dis-
aster management, mapping of buildings and their heights, simulation of
new buildings, updating and keeping cadastral data, change detection and
virtual reality. Urban, city, or town planning is the discipline of land use
planning which explores several aspects of the built and social environments
of municipalities and communities. The urban areas in the developing world
are under constant pressure of a growing population.[9] Efficient urban in-
formation system is a vital pre-requisite for planned development. It is
thought that LiDAR application to urbanization plans will bring to a more
sustainable and efficient growth of a city. In traffic line design area, LiDAR
technology offers a high precision DEM (Digital Elevation Model) for road
and railway design, to facilitate route design and the calculation of earth
work quantity. Besides, it is also of great application significance in route
design of communication network, oil pipe, air pipe. In digital city project,
a ”digital city” system can be established and perfected by combining the
high-precision DEM and GPS of LiDAR, and the data can be updated real
time therewith.[11] It is expected that LiDAR systems will be massively
used to plan the development of the cities of the future, to integrate new
infrastructure in existing contexts, taking into account sustainability and
real constraints due to environment, population and other key factors. In
other words, the rational urban densification and optimization, as well as
the development and construction of new buildings, streets, subway and
tram lines and innovative means of transportation is strictly related to the
implementation of this technology.

2.2.5 Energy
combining LiDAR technology and GIS (Geographic Information System) is
possible to develop a method for predicting city-wide electricity gains from
photovoltaic panels based on detailed 3D urban massing models combined
with Daysim-based hourly irradiation simulations, typical meteorological
year climactic data and hourly calculated rooftop temperatures. The re-
sulting data can be combined with online mapping technologies and search
engines as well as a financial module that provides building owners inter-
ested in installing a photovoltaic system on their rooftop with meaningful
data regarding spatial placement and informations.[10] So, in this sense Li-
2.3. STATE OF THE ART LIDAR SCANNING TECHNOLOGY 15

DAR could help the diffusion of RES (Renewable energy sources) in a more
efficient and rational way. This is only a simple example of what is possible
to conceive in the energy world thanks to LiDAR technology. An other im-
portant and relevant example is related to power transmission lines status
monitoring. In electric power transmission area, it can make for the laying,
maintenance and management of power grid. In the link of electric power
line design, the terrains and geographic features within the entire line design
area can be learned about based on LiDAR data. Especially in areas with
thick trees, the area and amount of trees to be cut down can be estimated.
In the link of power lines repair and maintenance, the height of line at any
location can be measured and calculated based on the LiDAR data points
on the electric power line and the elevation of corresponding exposed spot
on the ground, which can facilitate repair and maintenance.[11] Not only
Other relevant applications are missing for a time issues such as agriculture,
archeology or biology. Who is writing doesn’t want to explain in detail these
applications.

2.3 State of the art LiDAR scanning Technology

So, as said, this section is dedicated to the prior art. In other words the
understanding of the most important players on the LiDAR market at the
moment, also called incumbent, and a focus on the prominent players or new
comers.

• The world’s fourth largest tires manufacturer, Continental AG is a

leading German automotive manufacturing specialized in braking sys-
tems, automotive safety, tachographs etc. In 2016 was confirmed the
acquisition of the Hi-Res 3D Flash LiDAR business from Advanced
Scientific Concepts, Inc. (ASC) based in Santa Barbara, California.
This technology has further enhanced the company’s Advanced Driver
Assistance Systems product portfolio with a future-orientated solution
to add to the group of surrounding sensors needed to achieve highly
and fully automated driving. Nowadays Continental AG is one of the
most relevant players on the LiDAR market. A complete 3D model
of the vehicle surroundings nearby or over 200 meters away and as
close as few centimetres, is constructed in just 1.32 microseconds, 30
times per second. In addition, the distance to individual objects is also
accurately measured. The sensor is completely solid state, so no ro-
tating parts. The low complexity and high industrial feasibility mean
that it’s possible to efficiently install multiple sensors all around the
vehicle, thereby enabling the complete generation, real-time, 360 de-
grees images of the vehicle surroundings. The technology, has already
been deployed for space operations, provides a significantly more com-
prehensive and detailed 3D view of the entire vehicle surroundings –
16 CHAPTER 2. INTRODUCTION ON LIDAR SYSTEMS

both during the day and at night – and works reliably even in adverse
weather conditions. Mass production will start at the end of 2019 and
the product will be on the market in the early 2020, as scheduled.
The price is not known yet, but rumors say around 7000 dollars (even
though we don’t know if this could be a realistic number).

Figure 2.2: 3D flash LIDAR by Continental AG [13]

• Cosworth, the leading designer and manufacturer of high performance

engines and end-to-end automotive technologies, demonstrated a pro-
totype of solid state LiDAR sensor technology as part of the official
opening of its new state-of-the-art manufacturing center and North
American Headquarters in Shelby Township, last June.[14] Media and
invited guests witnessed a demonstration of a solid state LiDAR tech-
nology. Having no moving parts, the technology is designed and built
to provide stability that will meet the stringent automotive standards
for reliability. At the grand opening of North American manufac-
turing center they demonstrated the stability, precision, distance and
field of view that are necessary to make autonomous or assisted driving
safe. With a minimum of 3,200 lasers, a clear 120-degree filed of view
and 200 meter range, it is believed that this solution could be a game
changer for automotive sensor technology. Right now we have no more
information related to the product, large scale industrial production
and especially the price the LiDAR will be sold on the market.[14]
Anyway, since the author of this work consider that manufacturer and
product interesting it has been reported here. More information in the
following months.
2.3. STATE OF THE ART LIDAR SCANNING TECHNOLOGY 17

Description Value
Wavelenght 830 nm
Maximum pulse energy 5 nJ
Pulse Duration 4 ns
Pulse repetition frequency 2.16 MHz
Beam divergence (FWHM, full angle) 0.4 mrad
Mirror rotation 30 Hz
Base rotation 2.5 mHz

Table 2.1: BLK360 Laser Specs

• In this section is presented one of the LiDAR based products which I

believe is extremely interesting since its aim is similar to the one that
we have developed (later in this work we will talk about that exten-
sively). In detail we are talking about LEICA BLK360 which is used
to digitalize interiors and recreate the respective 3D representation.
The Leica BLK360 captures the world around you with full-colour
panoramic images overlaid on a high-accuracy point cloud. Simple to
use with just the single push of one button, the BLK360 is the small-
est and lightest of its kind. Anyone who can operate an iPad can now
capture the world around them with high resolution 3D panoramic
images. Using the ReCap Pro mobile app, the BLK360 streams image
and point cloud data to iPad. The app filters and registers scan data in
real time. After capture, ReCap Pro enables point cloud data transfer
to a number of CAD, BIM, VR and AR applications. The integration
of BLK360 and Autodesk software dramatically streamlines the real-
ity capture process thereby opening this technology to non-surveying
individuals. In few points:

– Allows you to scan in high, standard and fast resolutions;

– Weighs 1kg / Size 165 mm tall x 100 mm diameter;
– Less than 3 minutes for full-dome scan (in standard resolution)
and 150 MP spherical image generation;
– 360,000 laser scan pts/se
– HDR and thermal imaging

Table 2.1 reports BLK360 Laser propriety.

The laser incorporated in the product produces an invisible beam which

emerges from the rotating mirror. The laser product described in this section
is classified as laser class 1 in accordance with IEC prescriptions. This
18 CHAPTER 2. INTRODUCTION ON LIDAR SYSTEMS

Figure 2.3: Leica BLK360

products is safe under reasonably foreseeable conditions of operation. The

price for this product is quite prohibitive, around 16000$. Leica BLK360
can offer amazing performances and results but the price it’s a great obstacle
to its mainstream diffusion.
So, this is a brief presentation of the products I believe are more promis-
ing and interesting on the market. As I have said before the one we are more
interested in is Leica BLK360, because as specifications and characteristics
of the products is the closest to the LiDAR platform we have developed.
In fact we focused our attention on interior scanning and 3D environment
reconstruction. For sure our platform’s performance will be not comparable
with an industrial product which costs several thousands of dollars. Just to
be clear our budget is around 250$. Anyway, we consider the low budget a
positive aspect, in fact we want to show that is possible to get a good result
(high density point cloud) even though a less powerful hardware is used. In
this case it will be necessary to optimize and deploy at its best the hardware
we have in our hands. One of the biggest differences between our version
and Leica’s version will be for sure the acquisition speed: actually, it is said
that the commercial product is able to get a 360 degrees representation of
the environment roughly in 3 minutes. Probably, our prototype could scan
the environment for sure at a slower speed, maybe 6 times slower. We be-
lieve, though, is still ok, considering that time is not the principal driver for
our project.
2.4. LIDAR PROJECT SCOPE 19

2.4 LiDAR Project Scope

In this section a deeper focus on the description of the LiDAR platform’s
scope is given to the reader. A partial portrait or canvas has been presented
in §Chap. 1, §Sec. 1.2. Right now, we want to underline that the prototype
under the magnifying glass will provide:
• high resolution & high density Point Cloud;
• to each point will be assigned a colour, considering the distance from
the sensor;
• a navigation system based on Java which will allow the user to see in
real time the scanning result and translate and rotate the Point Cloud.
As we have already mentioned, at the moment the scanning speed it’s not
a big issue. At first our attention will be totally focused on getting a high
density and high resolution Point Cloud. Only, then we will change our fo-
cus on getting a higher scanning speed. The development of the final version
is a complex problem, achievable only through incremental steps. First of
all, it’s necessary to get familiar with the concept of LiDAR systems, how
do they work, their specifications, the safety rules needed to operate these
systems. Then we will move towards the first prototype which is a simple
LiDAR platform, more similar to a turret able to roughly scan the 3D space
which surrounds us, without any pretension to be accurate or detailed. The
aim of this first step is to get and accumulate experience, show to be familiar
with MCU control and writing algorithm able to execute what is intended.
In this first operative phase is necessary to understand that even a small
project is constitute by different sub-projects and it’s highly recommended
to schedule precisely the different parts and then assemble meticulously the
whole system solving in a smart way the integration problems. Only then it
will be possible to extensively develop the more advance and final version.
Starting from the limits and the problems emerged during phase 1, it will
be possible to study and conceive engineering solutions able to push forward
the performances of the machine and eventually get a better, more realistic
and industrial exploitable result.
Last but not the least during the whole process in essential to test small
parts of the platform in the Lab in order to validate components specifica-
tions, new possible solutions or the real boundaries of the platform itself.
So the engineering development is strictly linked to the test phase and the
pragmatic approach to problems.
In the next Chapter will be present the basic knowledge, necessary to un-
derstand how the following prototypes will work. We will pay attention to
details and protocols, as for example how the data transmission between
sensor and MCU works and other important topics. Eventually, then, it will
be possible to present how the LiDAR platforms has been made.
20 CHAPTER 2. INTRODUCTION ON LIDAR SYSTEMS
Chapter 3

Fundamentals on LiDAR

As we have said in the previous sections, especially the realization of this first
platform is essential to get confident with the equipment, and in particular
study more deeply the interaction between hardware and software. The
language necessary to control the MCU or better the micro controller that
we’ll use constitutes a completely new barrier. So, for this first period it’s
necessary to study and practice what’s written on books, web manuals and
tutorials.
In this Chapter will be reported and explained basic principles and how
things work; in particular:

• Sensor;

• Micro-Controller (Arduino);

• Servo Motors;

• Processing v3;

3.1 The Sensor

When space and weight requirements are tight, the LiDAR-Lite v3 soars.
It’s the ideal compact, high-performance optical distant measurement sensor
solution for drone, robot or unmanned vehicle applications. Using a single
chip signal processing solution along with minimal hardware, this highly
configurable sensor can be used as the basic building block for applications
where small size, light weight, low power consumption and high performance
are important factors. Featuring all of the core features of the popular
LiDAR-Lite v2, this easy-to-use 40 meter laser-based sensor uses about 130
milliamps during an acquisition. It is user-configurable so you can adjust
between accuracy, operating range and measurement time. It’s reliable,
powerful ranging and it’s a versatile proximity sensor. This is LIDAR-Lite
v3, a compact optical distance measurement sensor from Garmin. Each

21
22 CHAPTER 3. FUNDAMENTALS ON LIDAR

Specifications Measurements
Size (l x w x h) 20 x 48 x 40 mm
Weight 22 g
Operating Temperature -20 to 60 C

Table 3.1: Phisical Propriety Garmin Lidar Lite v3

Specifications Measurements
5 Vdc nominal
Power
4.5 Vdc min., 5.5 Vdc max.
105 mA idle
Current
135 mA continuous operations

Table 3.2: Electrical Propriety Garmin Lidar Lite v3

LiDAR-Lite v3 features an edge-emitting, 905nm (1.3 watts), single-stripe

laser transmitter, 4 m Radian x 2 m Radian beam divergence, and an optical
aperture of 12.5mm. The third version of the LiDAR-Lite still operates at
5V DC with a current consumption rate of <100mA at continuous operation.
On top of everything else, it can be interfaced via I2C or PWM with the
included 200mm accessory cable.
Note: CLASS 1 LASER PRODUCT CLASSIFIED EN/IEC 60825-1 2014.
This product is in conformity with performance standards for laser products
under 21 CFR 1040, except with respect to those characteristics authorized
by Variance Number FDA-2016-V-2943 effective September 27, 2016.

3.1.1 Technology and Operations

This device measures distance by calculating the time delay between the
transmission of a Near Infra-red laser and its reception after reflecting off a
target. This translates into distance using the known speed of light. The
unique signal processing approach transmits a coded signature and looks
for that signature in the return, which allows for highly reflective detection
with eye-safe laser power levels. Proprietary signal processing techniques are
used to achieve high sensitivity, speed and accuracy in a small, low power
and low-cost system.
To take a measurement, this device first performs a receiver bias correction
routine, correcting for changing ambient light levels and allowing maximum
sensitivity. Then the device sends a reference signal directly from the trans-
3.1. THE SENSOR 23

mitter to the receiver. It stores the transmit signature, sets the time delay
for ”zero” distance, and recalculates this delay periodically after several mea-
surements. Next, the device initiates a measurements by performing a series
of acquisitions. Each acquisition is a transmission of the main laser signal
while recording the return signal at the receiver. If there is a signal match,
the result is stored in memory as a correlation record. The next acquisition
is summed with the previous result. When an object at a certain distance
reflects the laser signal back to the device, these repeated acquisitions cause
a peak to emerge, out of the noise, at the corresponding distance location
in the correlation record.
The device integrates acquisitions until the signal peak in the correlation
record reaches a maximum value. If the returned signal is not strong enough
for this to occur, the device stops at the predetermined maximum acquisi-
tion count. Signal strength is calculated from the magnitude of the signal
record peak and a valid signal threshold is calculated from the noise floor.
If the peak is above the threshold the measurements is considered valid and
the device will calculate the distance, otherwise it will report 1 cm. When
beginning the next measurement, the device clears the signal and starts the
sequence again.[15]

3.1.2 Interface
Initialization On power-up or reset, the device performs a self-test se-
quence and initializes all registers with default values. After roughly 22 ms
distance measurements can be taken with the I2C interface or the Mode
Control Pin. So before explaining how LiDAR Lite v3’s I2C works it’s bet-
ter to give to the reader some basic and general informations related to I2C
itself. Only then it will be possible to understand clearly how the LiDAR
data transmission works.[15]

I2C description I2C-bus compatible ICs don’t only assist designers, they
also give a wide range of benefits to equipment manufacturers because:

• The simple 2-wire serial I2C-bus minimizes interconnections so ICs

have fewer pins and there are not so many PCB tracks; result - smaller
and less expensive PCBs;

• The completely integrated I2C-bus protocol eliminates the need for

address decoders and other ’glue logic’;

• The multi-master capability of the I2C-bus allows rapid testing and

alignment of end-user equipment via external connections to an assembly-
line;
24 CHAPTER 3. FUNDAMENTALS ON LIDAR

• The availability of I2C-bus compatible ICs in SO (small outline), VSO

(very small outline) as well as DIL packages reduces space requirements
even more.

These are just some of the benefits. In addition, I2C-bus compatible ICs in-
crease system design flexibility by allowing simple construction of equipment
variants and easy upgrading to keep designs up-to-date. In this way, an en-
tire family of equipment can be developed around a basic model. Upgrades
for new equipment, or enhanced-feature models (i.e. extended memory, re-
mote control, etc.) can then be produced simply by clipping the appropriate
ICs onto the bus. If a larger ROM is needed, it’s simply a matter of selecting
a micro-controller with a larger ROM. As new ICs supersede older ones, it’s
easy to add new features to equipment or to increase its performance by
simply unclipping the outdated IC from the bus and clipping on its succes-
sor.
For 8-bit oriented digital control applications, such as those requiring mi-
crocontrollers, certain design criteria can be established:

• A complete system usually consists of at least one microcontroller and

other peripheral devices such as memories and I/O expanders;

• The cost of connecting the various devices within the system must be
minimized;

• A system that performs a control function doesn’t require high-speed

data transfer;

• Overall efficiency depends on the devices chosen and the nature of the
interconnecting bus structure.

To produce a system to satisfy these criteria, a serial bus structure is needed.

Although serial buses don’t have the throughput capability of parallel buses,
they do require less wiring and fewer IC connecting pins. However, a bus is
not merely an interconnecting wire, it embodies all the formats and proce-
dures for communication within the system. Devices communicating with
each other on a serial bus must have some form of protocol which avoids
all possibilities of confusion, data loss and blockage of information. Fast
devices must be able to communicate with slow devices. The system must
not be dependent on the devices connected to it, otherwise modifications or
improvements would be impossible. A procedure has also to be devised to
decide which device will be in control of the bus and when. And, if different
devices with different clock speeds are connected to the bus, the bus clock
source must be defined. All these criteria are involved in the specification
of the I2C-bus.
The I2C-bus supports any IC fabrication process (NMOS, CMOS, bipolar).
Two wires, serial data (SDA) and serial clock (SCL), carry information
3.1. THE SENSOR 25

between the devices connected to the bus. Each device is recognized by

a unique address (whether it’s a microcontroller, LCD driver, memory or
keyboard interface) and can operate as either a transmitter or receiver, de-
pending on the function of the device. Obviously an LCD driver is only a
receiver, whereas a memory can both receive and transmit data. In addition
to transmitters and receivers, devices can also be considered as masters or
slaves when performing data transfers. A master is the device which initi-
ates a data transfer on the bus and generates the clock signals to permit
that transfer. At that time, any device addressed is considered a slave.[16]

I2C Lidar Lite Interface This device has a 2-wire, I2C-compatible serial
interface. It can be connected to an I2C bus as a slave device, under the
control of an I2C master device. It supports 400 kHz Fast Mode data trans-
fer. The I2C bus operates internally at 3.3 Vdc. An internal level shifter
allows the bus to run at the maximum of 5 Vdc. Internal 3k ohm pull-up
resistors ensure this functionality and allow for simple connection to the I2C
host. The device has a 7-bit slave address with the default value of 0x62.
The effective 8 bit-bit I2C address is 0x64 write and 0xC5. The device will
not respond to a general call. Support is not provided for 10-bit addressing.
The sensor module has a 7-bit slave address with a default value of 0x62 in
hexadecimal notation. The effective 8 bit I2C address is: 0x64 write, 0x65
read. The device will not presently respond to a general call. Please note
some additional informations:

• This device does not work with repeated START conditions. It must
first receive a STOP condition before a new START condition;

• The Ack and NACK item are responses from the master device to the
slave device;

• The last NACK in the read is technically optional, but the formal I2C
protocol states that the master shall not acknowledge the last byte.

So, the I2C serial bus protocol operates as follow:

• The master initiates data transfer by establishing a start condition,

which is when a high-to-low transition on the SDA line occurs while
SCL is high. The following byte is the address byte, which consists of
the 7-bit slave address followed by the read/write bit with zero state
indicating a write request. A write operation is used as the initial stage
of both read and write transfer. If the slave address corresponds to
the module’s address the unit responds by pulling SDA low during the
ninth clock pulse (this is termed the acknowledge bit). At this stage,
all other devices on the bus remain idle while the selected device waits
for data to be written to or read from its shift requester;
26 CHAPTER 3. FUNDAMENTALS ON LIDAR

Figure 3.1: LiDAR Lite I2C read/write operations

• Data is transmitted over the serial bus in sequences of ninth clock

pulses (eight data bits followed by an acknowledge bit). The transition
on the SDA line must occur during the low period of SCL and remain
stable during the high period of SCL;
• An 8 bit data byte following the address loads the I2C control the reg-
ister with the address of the control register to be read along with flags
indicating if auto increment of the addressed control register is desired
with successive reads or writes; and if access to the internal micro or
external correlation process register space is requested. Bit locations
5:0 contain the control register address while bit 7 enables the auto-
matic increment of control register with successive data blocks. Bit
position 6 selects correlation memory external to the micro-controller
if set. (Presently an advanced feature);
• If a read operation is requested, a stop bit is issued by the master at
the completion of the first data frame followed by the initiation of a
new start condition, slave address with the read bit set (one state).
The new address byte is followed by the reading of the one more data
bytes succession. After the slave has acknowledged receipt of a valid
address, data read operations proceed by the master releasing the I2C
3.2. ARDUINO 27

Specifications Measurements
Range (70% reflective Target) 40 m
Resolution ± 1 cm
Accuracy <5 m ± 2.5 cm
± 10 cm
Mean
Accuracy ≥ 5 m ± 1% of maximum distance
Ripple
± 1% of maximum distance
270 Hz typical
Update Rate (70% reflective Target) 650 Hz fast mode
>1000 Hz Short range only
50 Hz default
Repetition Rate
500 Hz max

Table 3.3: Garmin Lidar Lite v3 Performances

Specifications Measurements
Wavelength 905 nm (nominal)
Total Laser Power (peak) 1.3 W
Mode of operation Pulsed (256 pulsed max. pulse train)
Pulse width 0.5 µs (50% duty cycle)
Pulse train repetition frequency 10-20 kHz (nominal)
Energy per pulse < 280 nJ
Beam Diameter at laser aperture 12 x 2 mm
Divergence 8 m Radian

Table 3.4: Laser Propriety Garmin Lidar Lite v3

data line continuously clocking SCL. At the completion of the receipt

of a data byte, the master must strobe the acknowledge bit before
continuing the read cycle;
• For a write operation to proceed, Step 3 is followed by one or more 8 bit
data blocks with acknowledges provided by the slave at the completion
of each successful transfer. At the completion of the transfer cycle a
stop condition is issued by the master terminating operation.[15]

3.2 Arduino
Arduino Uno is a microcontroller board based on the ATmega328P. It has
14 digital input/output pins (of which 6 can be used as PWM outputs), 6
28 CHAPTER 3. FUNDAMENTALS ON LIDAR

Microcontroller ATmega328P
Operating Voltage 5V
Input Voltage (recommended) 7-12V
Input Voltage (limit) 6-20V
Digital I/O Pins 14 (of which 6 provide PWM output)
PWM Digital I/O Pins 6
Analog Input Pins 6
DC Current per I/O Pin 20 mA
DC Current for 3.3V Pin 50 mA
32 KB (ATmega328P)
Flash Memory
of which 0.5 KB used by bootloader
SRAM 2 KB (ATmega328P)
EEPROM 1 KB (ATmega328P)
Clock Speed 16 MHz
LED BUILTIN 13
Length 68.6 mm
Width 53.4 mm
Weight 25 g

Table 3.5: Arduino MCU specs

analog inputs, a 16 MHz quartz crystal, a USB connection, a power jack, an

ICSP header and a reset button. It contains everything needed to support
the microcontroller; simply connect it to a computer with a USB cable or
power it with a AC-to-DC adapter or battery to get started. It’s possible
to use Arduino UNO without worring too much about doing something
wrong, worst case scenario it’s needed to replace the chip for a few dollars
and start over again. ”Uno” means one in Italian and was chosen to mark
the release of Arduino Software (IDE) 1.0. The Uno board and version
1.0 of Arduino Software (IDE) were the reference versions of Arduino, now
evolved to newer releases. The Uno board is the first in a series of USB
Arduino boards, and the reference model for the Arduino platform; The
Arduino Uno has a resettable polyfuse that protects computer’s USB ports
from shorts and overcurrent. Although most computers provide their own
internal protection, the fuse provides an extra layer of protection. If more
than 500 mA is applied to the USB port, the fuse will automatically break
the connection until the short or overload is removed. The Uno differs from
all preceding boards in that it does not use the FTDI USB-to-serial driver
chip. Instead, it features the Atmega16U2 (Atmega8U2 up to version R2)
programmed as a USB-to-serial converter.
3.2. ARDUINO 29

3.2.1 Programming
Arduino Uno schematics is open-source hardware. The ATmega328 comes
preprogrammed with a bootloader that allows you to upload new code to
it without the use of an external hardware programmer. It communicates
using the original STK500 protocol. You can also bypass the bootloader
and program the microcontroller through the ICSP (In-Circuit Serial Pro-
gramming) header using Arduino ISP or similar; see these instructions for
details. The ATmega16U2 (or 8U2 in the rev1 and rev2 boards) firmware
source code is available in the Arduino repository. The ATmega16U2/8U2
is loaded with a DFU bootloader, which can be activated by:

• On Rev1 boards: connecting the solder jumper on the back of the

board (near the map of Italy) and then rese ing the 8U2;

• On Rev2 or later boards: there is a resistor that pulling the 8U2/16U2

HWB line to ground, making it easier to put into DFU mode;

You can then use Atmel’s FLIP software (Windows) or the DFU programmer
(Mac OS X and Linux) to load a new firmware. Or you can use the ISP
header with an external programmer (overwriting the DFU bootloader). See
this user-contributed tutorial for more information.
And then it’s possible to update and upload the code via USB thanks to
the appropriate software. It important to say that the arduino code is quite
simple and intuitive. To have an idea and understand how the basic Arduino
IDE environment is, please take a look at Fig. 3.2.

3.2.2 Connection Between Arduino and the Sensor

This example shows how to initialize, configure, and read distance from a
LIDAR-Lite connected over the I2C interface. Connections:

• LIDAR-Lite 5 Vdc (red) to Arduino 5v

• LIDAR-Lite I2C SCL (green) to Arduino SCL

• LIDAR-Lite I2C SDA (blue) to Arduino SDA

• LIDAR-Lite Ground (black) to Arduino GND

Somebody could raise the end and ask why it’s highly recommended to put
a capacitor between the ground and 5V pin. The answer is quite immediate
and easy: Capacitor is recommended to mitigate inrush current when device
is enabled and, as said before, and as it’s possible to see in Fig. 3.3
- 680uF capacitor (+) to Arduino 5v;
- 680uF capacitor (-) to Arduino GND;
30 CHAPTER 3. FUNDAMENTALS ON LIDAR

Figure 3.2: Arduino IDE

3.2.3 Power
The Arduino Uno board can be powered via USB connection or with an
external power supply. The power source is selected automatically. External
(non-USB) power can come either from an AC-to-DC adapter (wall-wart) or
battery. The adapter can be connected by plugging a 2.1mm center-positive
plug into the board’s power jack. Leads from a battery can be inserted
in the GND and Vin pin headers of the POWER connector. The board
can operate on an external supply from 6 to 20 volts. If supplied with less
than 7V, however, the 5V pin may supply less than five volts and the board
may become unstable. If using more than 12V, the voltage regulator may
overheat and damage the board. The recommended range is 7 to 12 volts.
The power pins are as follows (recap):

• Vin. The input voltage to the Arduino/Genuino board when it’s using
an external power source (as opposed to 5 volts from the USB connec-
tion or other regulated power source). You can supply voltage through
this pin, or, if supplying voltage via the power jack, access it through
this pin;

• 5V.This pin outputs a regulated 5V from the regulator on the board.

The board can be supplied with power either from the DC power jack
3.2. ARDUINO 31

Figure 3.3: Aduino-LiDAR Connection

Item Description Notes

you must observe the correct polarity
1 680 µF electrolytic capacitor
when installing the capacitor.
2 Power Ground (-) connection Black Wire
3 I2C SDA connection Blu wire
4 I2C SCL connection Green wire
Red wire
5 5 Vdc power (+) connection The sensor operates at 4.75 through 5.5 Vdc,
with a max. of 6 Vdc.

Table 3.6: Arduino-LiDAR Connection

(7 - 12V), the USB connector (5V), or the VIN pin of the board (7-
12V). Supplying voltage via the 5V or 3.3V pins bypasses the regulator,
and can damage your board;

• 3V3. A 3.3 volt supply generated by the on-board regulator. Maxi-

mum current draw is 50 mA;

• GND. Ground pins;

• IOREF. This pin on the Arduino/Genuino board provides the voltage

reference with which the microcontroller operates. A properly config-
ured shield can read the IOREF pin voltage and select the appropriate
power source or enable voltage translators on the outputs to work with
the 5V or 3.3V.
32 CHAPTER 3. FUNDAMENTALS ON LIDAR

3.2.4 Input and Outputs

See the mapping between Arduino pins and ATmega328P ports. The map-
ping for the Atmega8, 168, and 328 is identical.

Figure 3.4: Arduino ATMega Chipset

Each of the 14 digital pins on the Uno can be used as an input or output,
using pinMode(),digitalWrite(), and digitalRead() functions. They operate
at 5 volts. Each pin can provide or receive 20 mA as recommended operating
condition and has an internal pull-up resistor (disconnected by default) of
20-50kΩ. A maximum of 40mA is the value that must not be exceeded on
any I/O pin to avoid permanent damage to the micro-controller. In addition,
some pins have specialized functions:

• Serial: 0 (RX) and 1 (TX). Used to receive (RX) and transmit (TX)
TTL serial data. These pins are connected to the corresponding pins
of the ATmega8U2 USB-to-TTL Serial chip;

• External Interrupts: 2 and 3. These pins can be configured to trigger

an interrupt on a low value, a rising or falling edge, or a change in
value. See the attachInterrupt() function for details;

• PWM: 3, 5, 6, 9, 10, and 11. Provide 8-bit PWM output with the
analogWrite() function;

• SPI: 10 (SS), 11 (MOSI), 12 (MISO), 13 (SCK). These pins support

SPI communication using the SPI library;

• LED: 13. There is a built-in LED driven by digital pin 13. When the
pin is HIGH value, the LED is on, when the pin is LOW, it’s off;
3.2. ARDUINO 33

• TWI: A4 or SDA pin and A5 or SCL pin. Support TWI communica-

tion using the Wire library.

Communication Arduino/Genuino Uno has a number of facilities for

communicating with a computer, another Arduino/Genuino board, or other
micro-controllers. The ATmega328 provides UART TTL (5V) serial com-
munication, which is available on digital pins 0 (RX) and 1 (TX). An AT-
mega16U2 on the board channels this serial communication over USB and
appears as a virtual com port to software on the computer. The 16U2
firmware uses the standard USB COM drivers, and no external driver is
needed. However, on Windows, a .inf file is required. The Arduino Software
(IDE) includes a serial monitor which allows simple textual data to be sent
to and from the board. The RX and TX LEDs on the board will flash when
data is being transmitted via the USB-to-serial chip and USB connection to
the computer (but not for serial communication on pins 0 and 1). A Soft-
wareSerial library allows serial communication on any of the Uno’s digital
pins. The ATmega328 also supports I2C (TWI) and SPI communication.
The Arduino Software (IDE) includes a Wire library to simplify use of the
I2C bus; see the documentation for details. For SPI communication, use the
SPI library. Automatic (Software) Reset Rather than requiring a physical
press of the reset button before an upload, the Arduino/Genuino Uno board
is designed in a way that allows it to be reset by software running on a
connected computer. One of the hardware flow control lines (DTR) of the
ATmega8U2/16U2 is connected to the reset line of the ATmega328 via a
100 nanofarad capacitor. When this line is asserted (taken low), the reset
line drops long enough to reset the chip. The Arduino Software (IDE) uses
this capability to allow you to upload code by simply pressing the upload
button in the interface toolbar. This means that the bootloader can have a
shorter timeout, as the lowering of DTR can be well-coordinated with the
start of the upload. This setup has other implications. When the Uno is
connected to either a computer running Mac OS X or Linux, it resets each
time a connection is made to it from software (via USB). For the following
half-second or so, the bootloader is running on the Uno. While it is pro-
grammed to ignore malformed data (i.e. anything besides an upload of new
code), it will intercept the first few bytes of data sent to the board after a
connection is opened. If a sketch running on the board receives one-time
configuration or other data when it first starts, make sure that the software
with which it communicates waits a second after opening the connection
and before sending this data. The Uno board contains a trace that can be
cut to disable the auto-reset. The pads on either side of the trace can be
soldered together to re-enable it. It’s labeled ”RESET-EN”. You may also
be able to disable the auto-reset by connecting a 110 ohm resistor from 5V
to the reset line; see this forum thread for details.
34 CHAPTER 3. FUNDAMENTALS ON LIDAR

3.3 Servo Motors

Servo motors are specially designed motors to be used in control applica-
tions and robotics. They are used for precise position and speed control at
high torques. That’s why we thought that the servo could have been the
best choice for the LiDAR project. It consists of a suitable motor, position
sensor and a sophisticated controller. Servo motors can be characterized
according the motor controlled by servomechanism. Servo motors are avail-
able in power ratings from fraction of watt up to few 100 watts. They are
having high torque capabilities. The rotor of servo motor is made smaller
in diameter and longer in length, so that it has low inertia. What Is Ser-
vomechanism? Servomechanism is basically a closed-loop system Fig. 3.5,
consisting of a controlled device, controller, output sensor and feedback sys-
tem. The term servomechanism most probably applies to the systems where
position and speed is to be controlled. Mechanical position of the shaft can

Figure 3.5: Servo Motor, scheme

be sensed by using a potentiometer, which is coupled with the motor shaft

through gears. The current position of the shaft is converted into electrical
signal by the potentiometer, and then compared with the command input
signal. In modern servo motors, electronic encoders or sensors are used to
sense the position of the shaft. Command input is given according to the
required position of the shaft. If the feedback signal differs from the given
input, an error signal is generated. This error signal is then amplified and
applied as the input to the motor, which causes the motor to rotate. And
when the shaft reaches the required position, error signal becomes zero, and
hence the motor stays standstill holding the position.

3.3.1 How It Works

The simplicity of a servo is among the features that make them so reliable.
The heart of a servo is a small direct current (DC) motor, similar to what
you might find in an inexpensive toy. These motors run on electricity from
3.3. SERVO MOTORS 35

(a) Plastic Gears (b) Metal Gears

Figure 3.6: Stepper Gear examples

a battery or proper DC source and spin at high RPM (rotations per minute)
but put out very low torque. An arrangement of gears takes the high speed
of the motor and slows it down while at the same time increasing the torque.
It’s common sense the basic law of physics: work = f orce × distance. A
tiny electric motor does not have much torque, but it can spin really fast
(small force, big distance). The gear design inside the servo case converts
the output to a much slower rotation speed but with more torque (big force,
little distance). Even though, the amount of actual work is the same. Gears
in an inexpensive servo motor are generally made of plastic to keep it lighter
and less costly (see Fig. 3.6a). On a servo designed to provide more torque
for heavier work, the gears are made of metal (see Fig. 3.6b) and are harder
to damage. With a small DC motor, you apply power from a battery, and
the motor spins. Unlike a simple DC motor, however, a servo’s spinning
motor shaft is slowed way down with gears. A positional sensor on the
final gear is connected to a small circuit board (see Fig. 3.7). The sensor
tells this circuit board how far the servo output shaft has rotated. The
electronic input signal from the computer feeds into that circuit board. The
electronics on the circuit board decode the signals to determine how far the
user wants the servo to rotate. It then compares the desired position to the
actual position and decides which direction to rotate the shaft so it gets to
the desired position. That is what makes servo motors so useful: once you
tell them what you want done, they do the job without your help. This
automatic seeking behaviour of servo motors makes them perfect for many
robotic applications.
A lot of different kinds of servo motors are these days available:
• Positional rotation servo, this is the most common type of servo mo-
tor. The output shaft rotates in about 90 degrees, 180 degrees or 270
degrees. It has physical stops placed in the gear mechanism to prevent
turning beyond these limits to protect the rotational sensor;
• Continuous rotation servo, this is quite similar to the common po-
36 CHAPTER 3. FUNDAMENTALS ON LIDAR

Figure 3.7: Servo sensor

sitional rotation servo motor, except it can turn in either direction

indefinitely. The control signal, rather than setting the static posi-
tion of the servo, is interpreted as the direction and speed of rotation.
The range of possible commands causes the servo to rotate clockwise
or counter-clockwise as desired, at varying speed, depending on the
command signal;

• Linear servo, this is also like the positional rotation servo motor de-
scribed above, but with additional gears (usually a rack and pinion
mechanism) to change the output from circular to back-and-forth.
These servos are not easy to find though.

3.3.2 Control Strategy

Usually, the servo is controlled thanks to a micro-controller. In fact, the
motor presents 3 wires: GND, 5V and signal or better pulse (most of the
time, respectively, black or brown, red and orange or yellow). Intuitively, the
two wires responsible to power the motor must be connected to the respective
board pin of the controller, while the signal wire must be connected to a
PWM output pin.
Servos are controlled by sending an electrical pulse of variable width, or pulse
width modulation (PWM), through the control wire. There is a minimum
pulse, a maximum pulse, and a repetition rate. A servo motor can usually
only turn 90◦ or 135◦ in either direction for a total of 180◦ or 270◦ movement.
The motor’s neutral position is defined as the position where the servo has
the same amount of potential rotation in the both the clockwise or counter-
clockwise direction. The PWM sent to the motor determines the position of
3.4. PROCESSING V3 37

the shaft, and based on the duration of the pulse sent via the control wire;
the rotor will turn to the desired position. The servo motor expects to see a
pulse every 20 milliseconds (ms) or in other words 50Hz and the length of the
pulse will determine how far the motor turns. For example for a common
180◦ servo, a 1.5ms pulse will make the motor turn to the 90◦ position.
Shorter than 1.5ms moves it in the counter-clockwise direction toward the
0◦ position, and any longer than 1.5ms will turn the servo in a clockwise
direction toward the 180◦ position. The motor’s speed, then, is proportional
to the difference between its actual position and desired position. So if the
motor is near the desired position, it will turn slowly, otherwise it will turn
fast. This is called proportional control. This means the motor will only run
as hard as necessary to accomplish the task at hand, a very efficient way to
achieve the rotation, see Fig. 3.8.

Figure 3.8: Servo Control Strategy

In the next chapter will be presented the result of the servo test that has
been carried on in lab thanks to an oscilloscope (See §Chap. 4, §Sec. 4.2.1).

3.4 Processing v3
Processing is a flexible software sketchbook and a language for learning how
to code within the context of the visual arts. Since 2001, Processing has
promoted software literacy within the visual arts and visual literacy within
technology. There are tens of thousands of students, artists, designers, re-
searchers, and hobbyists who use Processing for learning and prototyping.
Its features are:[18]
• Free to download and open source;

• Interactive programs with 2D, 3D, PDF, or SVG output;

• OpenGL integration for accelerated 2D and 3D;

38 CHAPTER 3. FUNDAMENTALS ON LIDAR

• For GNU/Linux, Mac OS X, Windows, Android, and ARM;

• Over 100 libraries extend the core software;

• Well documented, with many books available;

Processing is an open-source graphical library and integrated development

environment (IDE) / playground built for the electronic arts, new me-
dia art, and visual design communities with the purpose of teaching non-
programmers the fundamentals of computer programming in a visual con-
text. Processing uses the Java language, with additional simplifications such
as additional classes and aliased mathematical functions and operations. As
well as this, it also has a graphical user interface for simplyfing the com-
pilation and execution stage. The Processing language and IDE were the
precursor to numerous other projects, notably Arduino, Wiring and P5.js.
Processing includes a sketchbook, a minimal alternative to an integrated
development environment (IDE)(See Fig. 3.9) for organizing projects.[19]
Every Processing sketch is actually a subclass of the PApplet Java class
(formerly a subclass of Java’s built-in Applet) which implements most of
the Processing language’s features.[20] When programming in Processing,
all additional classes defined will be treated as inner classes when the code
is translated into pure Java before compiling.[21] This means that the use
of static variables and methods in classes is prohibited unless Processing is
explicitly told to code in pure Java mode. Processing also allows for users
to create their own classes within the PApplet sketch. This allows for com-
plex data types that can include any number of arguments and avoids the
limitations of solely using standard data types such as: int (integer), char
(character), float (real number), and color (RGB, RGBA, hex). Processing
relates software concepts to principles of visual form, motion, and interac-
tion. It integrates a programming language, development environment, and
teaching methodology into a unified system. The Processing language is a
text programming language specifically designed to generate and modify im-
ages. Processing strives to achieve a balance between clarity and advanced
features. Beginners can write their own programs after only a few min-
utes of instruction, but more advanced users can employ and write libraries
with additional functions. The system facilitates teaching many computer
graphics and interaction techniques including vector/raster drawing, image
processing, color models, mouse and keyboard events, network communica-
tion, and object-oriented programming. Libraries easily extend Processing’s
ability to generate sound, send/receive data in diverse formats, and to im-
port/export 2D and 3D file formats.[22]
So, thanks to the concepts presented in this chapter it will be possible to
explain how the first LiDAR platform has been built.
3.4. PROCESSING V3 39

Figure 3.9: Processing v3 IDE

3.4.1 Export

The export feature packages a sketch to run within a Web browser. When
code is exported from Processing it is converted into Java code and then
compiled as a Java applet. When a project is exported, a series of files are
written to a folder named applet that is created within the sketch folder. All
files from the sketch folder are exported into a single Java Archive (JAR)
file with the same name as the sketch. For example, if the sketch is named
Sketch 123, the exported file will be called Sketch 123.jar. Every time a
sketch is exported, the contents of the applet folder are deleted and the files
are written from scratch. Any changes previously made to the index.html
file are lost. Media files not needed for the applet should be deleted from the
data folder before it is exported to keep the file size small. For example, if
there are unused images in the data folder, they will be added to the JAR file,
thus needlessly increasing its size. In addition to exporting Java applets for
the Web, Processing can also export Java applications for the Linux, Mac-
intosh, and Windows platforms. When ”Export Application” is selected
from the File menu, folders will be created for each of the operating sys-
tems specified in the Preferences. Each folder contains the application, the
source code for the sketch, and all required libraries for a specific platform.
Additional and updated information about the Processing environment is
available at www.processing.org/reference/environment or by selecting the
40 CHAPTER 3. FUNDAMENTALS ON LIDAR

”Environment” item from the Help menu of the Processing application.[22]

Chapter 4

LiDAR Platform v1

Firstly, it’s necessary to reaffirm that this first platform will be part of a
bigger learning phase. So, we will investigate how it’s possible to build
such a similar structure, discovering step by step problems and figuring out
how to overcome these issues. Moreover, this first version will be essential
to understand and define the boundaries of version one; at the same time
foresee a possible solution to improve the machine’s performance, towards
version 2.
Now, I will list the key components of the version that will be discussed in
detail in further sections of this chapter:

• Sensor;

• Servo motors (x 2);

• Arduino;

• Arduino Expansions;

• Simple Structure;

• Bluetooth, master and slave;

• Power Bank;

• Jumpers & cables.

As it’s possible to image, before putting all the parts together it’s necessary
to focus on the single parts, if required test in lab some specific components
and then assemble the whole platform solving integration problems or issues
and then test again. This last test is essential, since it’s like a benchmark
between the requirements on which the platform has been built and the
performances that the final product is actually able to deliver.
At the beginning of the process a general timeline has been deployed. The

41
42 CHAPTER 4. LIDAR PLATFORM V1

timeline itself presents different steps and phases, which correspond to differ-
ent implementations and functionalities. Going on with the process, taking
into account time constraint and possible improvements of the platform, we
will decide what to do as first or what to exclude. See Fig. 4.1.

Figure 4.1: Timeline detail v1

Anyway, right now it’s time to show what is the architecture of this first
prototype, see Fig. 4.2. It’s possible to see how the parts and tools have
been assembled together to form the LiDAR system. Since the structure
has to be portable, we have decided to provide the power necessary to feed
the equipment through a 5V output power bank. The full capacity of the
power supply will be 10000mAh, necessary to feed sensor, controller, motors
and the bluetooth. So, in simple words, the two motors will move the sensor
(pan and tilt movement), as soon as the movement is completed the sensor
acquisition is triggered by the Arduino, which contemporary knows and
control the servos, the measurement will be then elaborated by the MCU and
sent to the PC via Serial. Thanks to the adoption of the master and slave
wireless transmission of the informations, it’s possible to keep the circuit
simple, reduce the number of wires that is used without compromising the
transmission speed. So, the master device will be connected to the Arduino
Board expansion (directly mounted on the Arduino PCB itself) with four
pins: two for power (GND and 5V) and other two for data communication
(Tx and Sx). Obviously, then, the master will be coupled with his slave and
that one will be directly plugged into the USB port of the PC. In this way,
the slave could be powered and at the same time transmit data via Serial
Port to the Processing v3 Sketch that runs on the machine.
4.1. SENSOR CONTROL STRATEGY 43

The only thing that is missing, though it can be seen on the schematic, is
the presence of the Arduino servo expansion. By using that, we achieve: an
independence of the motors power supply for the Arduino and at the same
time the possibility of an easy control and feeding of the two servos.

Figure 4.2: LiDAR v1 Architecture

4.1 Sensor control Strategy

The Sensor control strategy will influence the whole project. The idea is to
control the sensor as a single shot gun: in fact, knowing the exact position of
the sensor in the space will, then, allow the triggering of the measurement.
When the process is completed the sensor will be then moved to the further
position and triggered again; and so on.
In this way the sensor will be used and exploited as a point-to-point acqui-
sition system. This control mode, seems to be the best taking into account
the features of the sensor itself and the hardware that will be coupled.
But one thing for sure it’s very important: how to use the sensor combined
to the Arduino Board. We have previously seen how to connect physically
the two entities. But, what about the software interface? Luckily, Garmin
(the sensor manufacturer) has released an Arduino library that makes easier
and intuitive how to control the LiDAR. In particular, it’s for sure firstly
 44 CHAPTER 4. LIDAR PLATFORM V1

necessary to install the library in the Arduino library folder. Once that this
step has been done, the sensor can be easily controlled and trigger, as said
before.
Firstly, it’s mandatory to define the LiDAR OBJECT with a proper syntax.
Then, it’s necessary to initiate the communication via I2C (the transmis-
sion mode we have chosen). Then, with a single command is possible to
trigger the sensor and capture the measurement. Anyway, the sensor can
operate in different modes and with two levels of accuracy. Obviously, the
deal it’s a compromise between speed and reliability of the data gathered
by the sensor. The following listed code explains basically how to operate
the sensor:
1 /∗−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
2 This example shows how t o i n i t i a l i z e , c o n f i g u r e ,
3 and r e a d d i s t a n c e from a LIDAR−L i t e c o n n e c t e d o v e r I2C
interface .
4
5 Connections :
6 LIDAR−L i t e 5 Vdc ( r e d ) t o Arduino 5v
7 LIDAR−L i t e I2C SCL ( g r e e n ) t o Arduino SCL
8 LIDAR−L i t e I2C SDA ( b l u e ) t o Arduino SDA
9 LIDAR−L i t e Ground ( b l a c k ) t o Arduino GND
10
11 ( C a p a c i t o r recommended t o m i t i g a t e i n r u s h c u r r e n t when d e v i c e
i s enabled )
12 680uF c a p a c i t o r (+) t o Arduino 5v
13 680uF c a p a c i t o r ( −) t o Arduino GND
14 −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−∗/
15 #i n c l u d e <Wire . h> // Standard L i b r a r y
16 #i n c l u d e <LIDARLite . h> // LiDAR ’ s L i b r a r y
17
18 LIDARLite myLidarLite ; // LiDAR OBJECT DEFINITION
19
20 void setup ( )
21 {
22 S e r i a l . b e g i n ( 1 1 5 2 0 0 ) ; // I n i t i a l i z e s e r i a l c o n n e c t i o n t o
display distance readings
23 /∗
24 begin ( i n t c o n f i g u r a t i o n , bool f a s t i 2 c , char l i d a r l i t e A d d r e s s )
25 −> S t a r t s t h e s e n s o r and I2C .
26
27 Parameters
28 −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
29 configuration : Default 0.
30 S e l e c t s one o f s e v e r a l p r e s e t c o n f i g u r a t i o n s .
31 f a s t i 2 c : D e f a u l t 100 kHz . I2C b a s e f r e q u e n c y . ( I f t r u e I2C
f r e q u e n c y i s s e t t o 400kHz ) ∗/
32 myLidarLite . b e g i n ( 0 , t r u e ) ; // S e t c o n f i g u r a t i o n t o d e f a u l t
and I2C t o 400 kHz
33 /∗ c o n f i g u r e ( i n t c o n f i g u r a t i o n , c h a r l i d a r l i t e A d d r e s s )
34 S e l e c t s one o f s e v e r a l p r e s e t c o n f i g u r a t i o n s .
35 Parameters
 4.1. SENSOR CONTROL STRATEGY 45

36 −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
37 configuration : Default 0.
38 0 : D e f a u l t mode , b a l a n c e d p e r f o r m a n c e .
39 1 : S h o r t range , h i g h s p e e d . Uses 0 x1d maximum a c q u i s i t i o n
count .
40 2 : D e f a u l t range , h i g h e r s p e e d s h o r t r a n g e . Turns on q u i c k
t e r m i n a t i o n d e t e c t i o n f o r f a s t e r measurements a t s h o r t r a n g e (
with d e c r e a s e d a c c u r a c y ) .
41 3 : Maximum r a n g e . Uses 0 x f f maximum a c q u i s i t i o n count .
42 4 : High s e n s i t i v i t y d e t e c t i o n . O v e r r i d e s d e f a u l t v a l i d
measurement d e t e c t i o n a l g o r i t h m , and u s e s a t h r e s h o l d v a l u e
f o r h i g h s e n s i t i v i t y and n o i s e .
43 5 : Low s e n s i t i v i t y d e t e c t i o n . O v e r r i d e s d e f a u l t v a l i d
measurement d e t e c t i o n a l g o r i t h m , and u s e s a t h r e s h o l d v a l u e
f o r low s e n s i t i v i t y and n o i s e . ∗/
44 myLidarLite . c o n f i g u r e ( 0 ) ; // Change t h i s number t o t r y out
alternate configurations
45 }
46 void loop ( )
47 {
48 /∗ d i s t a n c e ( b o o l b i a s C o r r e c t i o n , c h a r l i d a r l i t e A d d r e s s )
49 Take a d i s t a n c e measurement and r e a d t h e r e s u l t .
50 Parameters
51 −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−
52 b i a s C o r r e c t i o n : D e f a u l t t r u e . Take a c q u i s i t i o n with r e c e i v e r
b i a s c o r r e c t i o n . I f s e t t o f a l s e measurements w i l l be f a s t e r .
R e c e i v e r b i a s c o r r e c t i o n must be p er fo rm e d p e r i o d i c a l l y . ( e .
g . 1 out o f e v e r y 100 \ r e a d i n g s ) . ∗/
53
54 // Take a measurement with r e c e i v e r b i a s c o r r e c t i o n and p r i n t
to s e r i a l terminal
55 S e r i a l . p r i n t l n ( myLidarLite . d i s t a n c e ( ) ) ;
56
57 // Take 99 measurements w i t h o u t r e c e i v e r b i a s c o r r e c t i o n and
p r i n t to s e r i a l terminal
58 f o r ( i n t i = 0 ; i < 9 9 ; i ++)
59 {
60 S e r i a l . p r i n t l n ( myLidarLite . d i s t a n c e ( f a l s e ) ) ;
61 }
62 }
Listing 4.1: Code Listing-sensor acquisition
Taking a look at the code listing 4.1; it’s possible to see that on:
• Line 18, LiDAR Object has been defined;
• Line 32, I2C communication starts, with default settings (value = 0)
and speed set to 400kHz (true);
• Line 44, set LiDAR operational mode (value = 0, balanced perfor-
mances);
• Line 55, acquisition trigger (no bias correction). It means that the
measure is faster but not super accurate; as we have said the measure
46 CHAPTER 4. LIDAR PLATFORM V1

is a compromise between acquisition speed and accuracy. So it’s rec-

ommended to re-calibrate the sensor every 100 points, to get a more
precise distance value.

4.2 Servo control Strategy

The servo control strategy it’s pretty straight forward. Again, Arduino helps
us since the servo library it’s available. As we have said, standard servos
allow the shaft to be positioned at various angles, usually between 0 and
180 degrees. The Servo library supports up to 12 motors on most Arduino
boards and 48 on the Arduino Mega. Once, the servo is powered and the
pulse wire is connected to a PWM Arduino output pin, it’s necessary to
declare two servo entities (pan and tilt motor, motor x and motor y). Then,
the game is quite simple: in fact it’s necessary to remember that the servo’s
shaft can assume each position between 0◦ and 180◦ . To change the shaft’s
position it’s just needed to use the library’s instruction write. While, if we
want to know the exact position of the servo we can use the instruction read.
Both, read and write, are used as attributes or methods applied to the servo
entities.

Servo Trajectory At this point it will be briefly discussed an hot topic,

that will be treated for sure with a deeper focus in following sections. The
final density of the point cloud1 is linked for sure to the reliability of the
sensor and its capability to trigger and gather distances in the smallest
possible period of time. But, assuming that the sensor is fast enough, as it
actually is, the main driver that determines the density of the point cloud
is the minimum step that the servo can execute. Taking a look on the
internet and LiDAR commercial products, it’s noticeable that extremely
detailed point clouds reach, at least, densities around 1 or 2 million points
(over a 3D sapace which is roughly 180◦ × 90◦ , in other words half of a
semi-sphere). For this first prototype, though, the density is absolutely
lower, around 200000 points. This comes from a very simple mathematical
calculation, considering the smallest servo’s movement equal to 1 degree
and a scanned 3D space which is actually close to half of a semi-sphere.
At this point, we don’t know yet if this density is enough to distinguish,
at least roughly, obstacles and scanned objects. In the final phase of the
development step this aspect will be checked and discussed: if the result
it’s not considered acceptable, the version2 will be completely focused on a
1
A point cloud is a set of data points in space. Point clouds are generally produced by
3D scanners, which measure a large number of points on the external surfaces of objects
around them. As the output of 3D scanning processes, point clouds are used for many
purposes, including to create 3D CAD models for manufactured parts, for metrology and
quality inspection, and for a multitude of visualization, animation, rendering and mass
customization applications.
4.2. SERVO CONTROL STRATEGY 47

solution able to increment dramatically the density and performances in

general. Unfortunately, at least for version1, the servo library allows a
minimum step of 1◦ , any smaller rotation is prohibited and impossible to be
executed.
Last but not the least, it’s necessary to talk a bout the trajectory that the
pan and tilt motor will describe: the pan motor will sweep the horizontal
plane (rotation around a vertical axis), while the tilt motor will sweep the
vertical plane (rotation around a horizontal axis). When the Arduino board
is turned on, the reset operation brings the servo at homing position. In
particular, the homing position is around the middle of the entire interval.
So, then, the servo will swipe more or less 85 degrees clockwise and 85
degrees anti-clockwise. Once, a complete 180◦ angle is described, the tilt
servo will move 1 degree upward.

4.2.1 Servo characteristic determination

At this point, though, we posed ourselves a question: ”How do we know that
the servo characteristic is linear?”. In other words, who is assuring us that
when we send the instruction servo.write(37) the signal sent through the
pulse wire corresponds to a precise PWM 50 Hz signal interpreted by the
servo as a rotation to position 37, for example. For this reason, we designed
a simple test in lab, able to define and give a precise idea of the servo
characteristic. By using a common digital oscilloscope synchronized to the
servo pulse wire and sending firstly the pulses from PC, thanks to a testing
tool provided with servos, it’s possible to check this out. The first thing that
we discovered was that the PWM train is actually 50Hz, as expected; but
then we noticed that for our servos, the width of the pulse varies between
500µs (which corresponds to 0◦ ) and 2500µs (which corresponds to 180◦ ).
Before taking a look at the data, it’s reasonable to explain better how the
measurement has been carried on. On the Pc has been installed a servo
test tool, specifically design to control the Arduino Servo expansion. This
tool, is able to produce a pulsed signal that drives the servos, through the
control board. The notable advantage is that it’s easily possible to drive the
motor thanks to the Pc interface. The general purpose of the experience
is to collect roughly 20 points and identify the characteristic of the servo
motor. By varying the position of the shaft on the Pc interface and then
measuring the PWM signal thanks to the oscilloscope, it’s possible to check
the amplitude of the signal, as well as the frequency. See Fig. 4.4. In the
picture, it’s possible to notice, on the right bottom part of the oscilloscope’s
screen, that the signal’s frequency it’s around 50Hz (precisely 49,951Hz) and
the width of the pulse corresponding to 180◦ is 2500µs, as expected. Anyway,
in the preliminary part of the lab experience it has been noticed that the
PWM signal near the boundaries (0 and 180) in quite unstable, so it’s been
decided not to drive the servo with angles smaller than 10◦ and not grater
48 CHAPTER 4. LIDAR PLATFORM V1

Figure 4.3: Servo lab test 1

than 170◦ , since there’s no linearity in this portion of the characteristic.

So, then, the characteristic has been drawn testing 20 positions, especially
close to the upper and lower limit of the servo previously fixed. Taking a
look at Tab. 4.1 and to Fig 4.5 it’s possible to see clearly that the servo
characteristic is pretty linear, or at least it’s quite possible to consider it
linear. This was the result we were looking for, considering the fact that
the resolution of the point cloud and the accuracy of the point positioning
is strictly dependent on the servos.

4.3 Data Elaboration and Transmission

Right now, this chapter has explained the main mechanism around the sen-
sor acquisition and servo controlling. The output of the LiDAR sensor mea-
surement is a distance, associated to two angular positions, respectively re-
lated to the two motors. It’s intuitive that the sensor describes a trajectory
that moves on a semi-sphere. At the same time, though, the final repre-
sentation that we want is a three dimensional picture based on a Cartesian
coordinate system.

T(ρ,θ,φ) ⇒ T(x,y,z) (4.1)

4.3. DATA ELABORATION AND TRANSMISSION 49

Degrees Pulse Amplitude (µs)

10 600
12 640,8
14 660,4
16 680,5
18 700,3
20 720,5
22 740,4
24 760,2
26 780,5
28 800,5
30 820,3
32 840
90 1502
150 2161
151 2180
153 2200
155 2220
157 2241
160 2280
163 2030
165 2320
169 2360
170 2400

Table 4.1: Servo Characteristic

50 CHAPTER 4. LIDAR PLATFORM V1

Figure 4.4: Servo lab test 2

In mathematics, a spherical coordinate system is a coordinate system for

three-dimensional space where the position of a point is specified by three
numbers: the radial distance of that point from a fixed origin, its polar
angle measured from a fixed zenith direction, and the azimuth angle of its
orthogonal projection on a reference plane that passes through the origin
and is orthogonal to the zenith, measured from a fixed reference direction
on that plane. It can be seen as the three-dimensional version of the polar
coordinate system. The radial distance is also called the radius or radial
coordinate. The polar angle may be called colatitude, zenith angle, normal
angle, or inclination angle. The use of symbols and the order of the co-
ordinates differs between sources. In one system frequently encountered in
physics (ρ, θ, φ) gives the radial distance, polar angle, and azimuthal angle,
whereas in another system used in many mathematics books (ρ, θ, φ) gives
the radial distance, azimuthal angle, and polar angle. Other conventions
are also used, so great care needs to be taken to check which one is being
used. A number of different spherical coordinate systems following other
conventions are used outside mathematics. In a geographical coordinate
system positions are measured in latitude, longitude and height or altitude.
There are a number of different celestial coordinate systems based on differ-
ent fundamental planes and with different terms for the various coordinates.
The spherical coordinate systems used in mathematics normally use radians
rather than degrees and measure the azimuthal angle counter-clockwise from
4.3. DATA ELABORATION AND TRANSMISSION 51

Figure 4.5: Servo Characteristic, graph

the x-axis to the y-axis rather than clockwise from north (0◦ ) to east (90◦ )
like the horizontal coordinate system. The polar angle is often replaced
by the elevation angle measured from the reference plane. Elevation angle
of zero is at the horizon. The spherical coordinate system generalizes the
two-dimensional polar coordinate system. It can also be extended to higher-
dimensional spaces and is then referred to as a hyper-spherical coordinate
system.
Spherical coordinates determine the position of a point in three-dimensional
space based on the distance ρ from the origin and two angles θ and φ. The
following graphics (See Fig. 4.6) may help to understand spherical coordi-
nates better. On this page, we derive the relationship between spherical and
Cartesian coordinates.

Relationship between spherical and Cartesian coordinates Spher-

ical coordinates are defined as indicated in the following figure, which il-
lustrates the spherical coordinates of the point P. The coordinate ρ is the
distance from P to the origin. If the point Q is the projection of P to
the xy-plane, then θ is the angle between the positive x-axis and the line
segment from the origin to Q. Lastly, φ is the angle between the positive
z-axis and the line segment from the origin to P. We can calculate the
relationship between the Cartesian coordinates (x, y, z) of the point P and
its spherical coordinates (ρ, θ, φ) using trigonometry. The pink triangle (See
Fig. 4.7) is the right triangle whose vertices are the origin, the point P, and
its projection onto the z-axis. As the length of the hypotenuse is ρ and φ
52 CHAPTER 4. LIDAR PLATFORM V1

Figure 4.6: Spherical Coordinates

Figure 4.7: Spherical & Cartesian Coordinates

is the angle the hypotenuse makes with the z-axis leg of the right triangle,
the z-coordinate of P (i.e., the height of the triangle) is z = ρ cos φ. The
length of the other leg of the right triangle is the distance from P to the
z-axis, which is r = ρ sin φ. The distance of the point Q from the origin is
the same quantity. The cyan triangle, shown in both the original 3D coordi-
nate system on the left and in the xy-plane on the right, is the right triangle
whose vertices are the origin, the point Q, and its projection onto the x-axis.
In the right plot, the distance from Q to the origin, which is the length of
hypotenuse of the right triangle, is labelled just as r. As θ is the angle this
hypotenuse makes with the x-axis, the x- and y-components of the point Q
(which are the same as the x- and y-components of the point P) are given
by x = r cos θ and y = r sin θ. Since r = ρsinθ, these components can be
rewritten as x = ρ sin φ cos θ and y = ρ sin φ sin θ. In summary, the formulas
 4.4. DATA PROCESSING, VISUALIZATION AND RESULTS 53

for Cartesian coordinates in terms of spherical coordinates are:

x = ρ sin φ cos θ
y = ρ sin φ sin θ
z = ρ cos θ

So, once the sensor has revealed the distance, there are some lines in the
Arduino code that operates the transformation and then print a string with
(x, y, z) on the serial port. This operation it’s quite easy thanks to the
instruction println. It is as well possible to print the data to file, but actually
we preferred to transmit the data via Serial: in this way it’s immediate a
real time elaboration.

4.4 Data Processing, Visualization and Results

Since the scanning process has been totally explained, now it’s the visual-
ization turn. Once, the data has been collected, then there’s the need to
show them and create a virtual environment which let us appreciate the
shapes of scanned objects. Moreover, as we have chosen, we want it to be
real time, which means that the scanning process and visualization process
are contemporary.
In other words, as previously hinted in §Chap. 3, §Sec. 3.4, a processing
sketch running on the Pc will show the point cloud. The sketch should:

• Read real time the Serial Port;

• Store temporarily the data in a matrix;

• Plot the data without delays;

• Assign a proper color to each point;

• Offer a proper navigation system;

• Allow to save data.

As a matter of facts, the Arduino output is (x, y, z) coordinates, so the

processing v3 sketch accepts (x, y, z) data as input. The big deal is to set
the correct Serial Port speed, synchronized to the Arduino transmission
baud rate.
1 // This s k e t c h a c c e p t s XYZ c o o r d i n a t e s from Arduino LIDAR
s c a n n e r and d i s p l a y s them g r a p h i c a l l y a s a 3D p o i n t c l o u d
t h a t you can pan , zoom , and r o t a t e u s i n g keyboard .
2
3 import p r o c e s s i n g . s e r i a l . ∗ ; // S e r i a l Port Java l i b r a r y
4 import j a v a . i o . F i l e W r i t e r ; // l i b r a r i e s n e c e s s a r y t o e x p o r t
file
 54 CHAPTER 4. LIDAR PLATFORM V1

5 import j a v a . u t i l . Calendar ;
6 import j a v a . t e x t . SimpleDateFormat ;
7 \\ g l o b a l v a r i a b l e i n i t i a l i z a t i o n
8 Serial serial ;
9 i n t serialPortNumber = 0 ;
10 f l o a t angle = 6.5 f ;
11 f l o a t angleIncrement = 0;
12 f l o a t xOffset = 3.0;
13 f l o a t xOffsetIncrement = 0;
14 f l o a t yOffset = 152.0 f ;
15 f l o a t yOffsetIncrement = 0;
16 f l o a t scaleIncrement = 0;
17 float h;
18 A r r a y L i s t <PVector> v e c t o r s ; \\ v e c t o r i n i t i a l i z a t i o n t o s t o r e
data
19 int lastPointIndex = 0;
20 int lastPointCount = 0;
21
22 void setup ( ) {
23 s i z e ( 8 0 0 , 6 0 0 , P3D) ; // environment i n i t i a l i z a t i o n
24 colorMode (HSB, 3 6 0 , 1 0 0 , 1 0 0 ) ;
25 noSmooth ( ) ;
26 v e c t o r s = new A r r a y L i s t <PVector >() ;
27 S t r i n g [ ] s e r i a l P o r t s = S e r i a l . l i s t ( ) ; // s e r i a l p o r t r e a d −>
28 String s e r i a l P o r t = s e r i a l P o r t s [ serialPortNumber ] ;
29 p r i n t l n ( ” Using s e r i a l p o r t \” ” + s e r i a l P o r t + ” \” ” ) ;
30 p r i n t l n ( ”To u s e a d i f f e r e n t s e r i a l port , change
serialPortNumber : ” ) ;
31 printArray ( s e r i a l P o r t s ) ;
32 s e r i a l = new S e r i a l ( t h i s , s e r i a l P o r t , 1 1 5 2 0 0 ) ; } // <−
33 v o i d draw ( ) {
34 String input = s e r i a l . readStringUntil (10) ;
35 i f ( i n p u t != n u l l ) {
36 S t r i n g [ ] components = s p l i t ( input , ’ ’ ) ;
37 i f ( components . l e n g t h == 3 ) {
38 v e c t o r s . add ( new PVector ( f l o a t ( components [ 0 ] ) , f l o a t (
components [ 1 ] ) , f l o a t ( components [ 2 ] ) ) ) ; } }
39 background ( 0 ) ;
40 t r a n s l a t e ( width / 2 , h e i g h t / 2 , −50) ;
41 rotateY ( angle ) ;
42 int size = vectors . size () ;
43 f o r ( i n t i n d e x = 0 ; i n d e x < s i z e ; i n d e x++) {
44 PVector v = v e c t o r s . g e t ( i n d e x ) ;
45 i f ( i n d e x == s i z e − 1 ) {
46 // draw r e d l i n e t o show r e c e n t l y added LIDAR s c a n p o i n t
47 i f ( i n d e x == l a s t P o i n t I n d e x ) {
48 l a s t P o i n t C o u n t ++;
49 } else {
50 lastPointIndex = index ;
51 lastPointCount = 0;
52 }
53 i f ( lastPointCount < 10) {
54 s t r o k e (0 , 100 , 100) ;
55 l i n e ( x O f f s e t , y O f f s e t , 0 , v . x ∗ s c a l e + x O f f s e t , −v . z ∗
 4.4. DATA PROCESSING, VISUALIZATION AND RESULTS 55

s c a l e + y O f f s e t , −v . y ∗ s c a l e ) ; } }
56 h= s q r t ( v . x∗v . x+v . y∗v . y+v . z ∗v . z ) ;
57 s t r o k e (360 −(h ∗ 0 . 4 ) , 1 0 0 , 1 0 0 ) ;
58 p o i n t ( v . x ∗ s c a l e + x O f f s e t , −v . z ∗ s c a l e + y O f f s e t , −v . y ∗
scale ) ;}
59 a n g l e += a n g l e I n c r e m e n t ;
60 x O f f s e t += x O f f s e t I n c r e m e n t ;
61 y O f f s e t += y O f f s e t I n c r e m e n t ;
62 s c a l e += s c a l e I n c r e m e n t ; }
63 void keyPressed ( ) {
64 i f ( key == ’ q ’ ) {
65 // zoom i n
66 scaleIncrement = 0.02 f ;
67 } e l s e i f ( key == ’ z ’ ) {
68 // zoom out
69 s c a l e I n c r e m e n t = −0.02 f ;
70 } e l s e i f ( key ==’ p ’ ) {
71 // e r a s e a l l p o i n t s
72 vectors . clear () ;
73 } e l s e i f ( key == ’ a ’ ) {
74 // move l e f t
75 x O f f s e t I n c r e m e n t = −1 f ;
76 } e l s e i f ( key == ’ d ’ ) {
77 // move r i g h t
78 xOffsetIncrement = 1 f ;
79 } e l s e i f ( key == ’w ’ ) {
80 // move up
81 y O f f s e t I n c r e m e n t = −1 f ;
82 } e l s e i f ( key == ’ x ’ ) {
83 // move down
84 yOffsetIncrement = 1 f ;
85 } e l s e i f ( key == CODED) {
86 i f ( keyCode == LEFT) {
87 // r o t a t e l e f t
88 a n g l e I n c r e m e n t = −0.015 f ;
89 } e l s e i f ( keyCode == RIGHT) {
90 // r o t a t e r i g h t
91 angleIncrement = 0.015 f ;}}}
92 void keyReleased ( ) {
93 i f ( key == ’ q ’ ) {
94 scaleIncrement = 0 f ;
95 } e l s e i f ( key == ’ z ’ ) {
96 scaleIncrement = 0 f ;
97 } e l s e i f ( key == ’ a ’ ) {
98 xOffsetIncrement = 0 f ;
99 } e l s e i f ( key == ’ d ’ ) {
100 xOffsetIncrement = 0 f ;
101 } e l s e i f ( key == ’w ’ ) {
102 yOffsetIncrement = 0 f ;
103 } e l s e i f ( key == ’ x ’ ) {
104 yOffsetIncrement = 0 f ;
105 } e l s e i f ( key == ’ s ’ ) {
106 saveToFile () ;
107 } e l s e i f ( key == CODED) {
 56 CHAPTER 4. LIDAR PLATFORM V1

108 i f ( keyCode == LEFT) {

109 angleIncrement = 0 f ;
110 } e l s e i f ( keyCode == RIGHT) {
111 angleIncrement = 0 f ;}}}
112 // Fu nction t o s a v e t h e p o i n t c l o u d i n t h e P r o c e s s i n g i n s t a l l
directory
113 void saveToFile ( ) {
114 S t r i n g f i l e N a m e = ” . / p o i n t s ”+
115 new SimpleDateFormat ( ”yyMMdd HHmmss” ) . for ma t ( Calendar .
g e t I n s t a n c e ( ) . getTime ( ) )+” . xyz ” ;
116 P r i n t W r i t e r pw = n u l l ;
117 try {
118 pw = new P r i n t W r i t e r ( new F i l e W r i t e r ( fileName , t r u e ) ) ;
119 f o r ( i n t i =0; i <v e c t o r s . s i z e ( ) ; i ++)
120 pw . p r i n t l n ( ( i n t ) v e c t o r s . g e t ( i ) . x + ” ” +
121 ( int ) vectors . get ( i ) . y + ” ” +
122 ( int ) vectors . get ( i ) . z ) ;}
123 catch ( Exception e ) {
124 }finally {
125 i f (pw != n u l l ) pw . c l o s e ( ) ; } }
Listing 4.2: Processing sketch

In the following part the algorithm will be commented to explain the differ-
ent parts and functionalities. As for each scratch or program at the begin-
ning are reported the libraries that are needed to run, compile and interpret
the algorithm. Then the global variables are exposed.

• Line 23, this instruction declares and initialize the size and the colour
of the environment trough which the point cloud will be showed. It will
be a rectangle 800 × 600 pixels. Before drawing 3D form in Processing,
it’s necessary to tell the software to draw with a 3D renderer. The
default renderer in Processing draws only two-dimensional shapes, but
there are additional options (P3D and OPENGL) to render 3D form.
P3D is the simplest and most compatible renderer, and it requires no
additional libraries;

• Line 24, definition of the color mode and the background colour of the
environment. Processing uses the RGB color model as its default for
working with color, but the HSB specification can be used instead to
define colors in terms of their hue, saturation, and brightness. The hue
of a color is what most people normally think of as the color name:
yellow, red, blue, orange, green, violet. A pure hue is an undiluted
color at its most intense. The saturation is the degree of purity in a
color. It is the continuum from the undiluted, pure hue to its most
diluted and dull. The brightness of a color is its relation to light and
dark.[22]
For example the color black is represented as (HSB, 360, 100, 100).
Somebody could raise the hand and ask why the system HSB has been
4.4. DATA PROCESSING, VISUALIZATION AND RESULTS 57

chosen rather the more common and intuitive RGB. The answer lies
on something that well be explained in detail later. In fact, a color
will be assigned to each point taking into account the distance from
the sensor. If the system RGB is adopted, the result is a saturation of
the color and points which are close to each other are displayed with
the same color. This is something that we don’t want since the color
to the points contributes to create a more realistic three-dimensional
effect;

• Line 26-32, initialization of the vector in which data are then stored.
Establishment of the Serial Port communication;

• Line 33, this part is responsible to draw the points, in the proper
position, according to the coordinates. In order to make it easier to
understand and intuitive, the last point that is displayed on the screen
is connected by means of a red line to the origin of the axis. In this
way, it’s possible to follow the progression in the representation of the
point cloud;

• Line 66-111, this part of the code is responsible of the navigation sys-
tem. By pressing the keyboard arrows is possible to rotate the point
cloud around a vertical or horizontal axis. While, pressing W,S,D,A
buttons is possible to translate the point cloud horizontally or verti-
cally. Finally, pressing the Q is possible to zoom in, while Z corre-
sponds to zoom out. Last but not least important, P erases all points
and S saves the point cloud to file;

• Line 113-125, is the part related to the function save to file. By press-
ing the S button, the function save is invoked. The file in output will
contain the three-dimensional coordinates, the file will be saved in Pro-
cessing v3 installation folder with a name ”yyMMdd HHmmss.xyz”.

Results In this paragraph are reported samples from the first test carried
on with LiDAR platform v1. As it’s easy to understand, at the very be-
ginning the processing sketch was not able to display colors (See Fig. 4.8
and Fig. 4.9). While Fig. 4.10 represent the evolution to colors. As, said
before, colors help to identify better objects and give a more accurate three-
dimensional aspect. Anyway, it’s necessary to be honest and say that this
first platform has given a result which is not good as expected. In fact,
the density of the point cloud is not sufficient to give us a precise idea of
the object or obstacle. The only thing that we can infer is the approximate
shape of things or rooms. This prototype just underlines and shows which
is possible to have a 3D representation of a surrounding environment em-
ploying a simple hardware as we did. But, for sure, the platform requires
improvements, as well as upgrades. The first noticeable thing is the point
58 CHAPTER 4. LIDAR PLATFORM V1

Figure 4.8: Scan 1

Figure 4.9: Scan 2

cloud density: it’s not enough to show with precision the scanned objects.
For this reason, the next version should improve dramatically this param-
eter. In the following section new proposal to solve this problem will be
presented.

4.5 Towards LiDAR v2

The first step, has been to find out the critical parts of our project, start-
ing from a free flow of thoughts or brainstorming. But, as well known, is
not possible to randomly presents the list that has emerged, so the second
step has been to sort that list. From the most critical to the less critical.
In this way is possible to establish a priority and fix these problems fol-
lowing a well-defined path and process. It’s not possible to think that the
4.5. TOWARDS LIDAR V2 59

Figure 4.10: Scan 3

Rank Priorities
Code – algorithm (Servo control, Serial Data
1
Transmission, Data Processing)
2 Motors
3 Structure

Table 4.2: Table of enhancement

60 CHAPTER 4. LIDAR PLATFORM V1

improvement of the project only relies on a better hardware. Rather than

that, the hardware upgrade must be followed by an improvement in the code
and the algorithm. So many times, the experience has showed us that the
hardware limits (constraints related to a limited amount of resources, such
as money and time) can be overcome by software manipulation and smart
solutions. Following this approach, we identified three fields of interventions:

1)

a. Servo control technique: our technique seems to be not so reliable

and it presents some unsolved problems and bugs (for example the
execution of an angle greater than 120◦ ). Or another issue is related
to the possibility to reduce the angle between two steps. Right now,
it’s approximately around 1◦ , but in following version we would like to
bring this value to 0.5◦ or smaller. This, could lead to a more efficient
and accurate scanning performance;

• + More efficient, accurate scanning technique;

• + It’s possible to get a 3D representation of small objects or
details;
• - Non linearities. When the step is smaller than 1◦ we’re not sure
about the real entity of the servo movement;
• - Lack of accuracy and reliability.

b. Serial COM: a reliable and affordable serial data transmission permits

a real data acquisition and the possibility to show a real time point
cloud. Right now, the serial communication presents some problems
and it’s not so stable. In fact, there are some errors in the data format,
so this implies a more difficult data elaboration and as consequence
the process will be more expensive in terms of time and less stable.

• + No dead time between data acquisition and processing;

• + Show Real Time point cloud;
• + More effective and efficient way to operate the scan.

c. Data processing: implement a better data elaboration means to have

a precise idea of what you’re scanning. So, a smart data processing
involves an advanced algorithm able to discern which points are useful
and which are not (errors) and showing them in an easy, immediate
and understandable way. Another issue which is close to that is to
assign to each point the right color based on the distance from the
detecting device.

• + Real Time Solution;

4.5. TOWARDS LIDAR V2 61

• + Easy to display.

All of those three options, software related, are not cost effective, since
there’s no need to buy anything but just programming and code. Moreover,
it’s true that on the internet it’s possible to find some open source software
that can guide you in the development of a new application or a new soft-
ware part. But at the same time, we should say that we should spend a
lot of time to make this process smooth and reliable. Anyway, regarding
the motors, it’s necessary to evaluate if a different technology could bring
to better performances.

2)

a. To reach a high level of accuracy it’s possible to buy new servos with
better performances and specs. The main problem that we must face
is the instability of the PWM signal to the servos. In fact, we have
tested the servo with an oscilloscope and we have found out that the
PWM signal is not stable and oscillates. This can lead to servo rota-
tions that are not totally controllable and at the same time uncertain.
Considering these problems, it’s a wise decision to take into account
different motor technologies, for example stepper motors. Stepper are
used in critical applications where the minimum required step is re-
ally small or in task where the precise positioning is an issue. With
an improvement of the motor performances it’s possible to scan more
accurately the surrounding environment.

• + Accuracy;
• + Reliability;
• + Increase the performance of the whole system;
• - Time and Cost, expensive (especially switch to other technolo-
gies);
• - Test in lab;
• - Slow down the scanning speed (↑ accuracy ↓ speed).

3)

a. The mechanical structure is very important because it has to prevent

the diffusion of vibrations and place the sensor in a precise position,
well fixed and still. The idea is to design a more robust and reliable,
stable and safe structure. One thing that we want to avoid is glued
parts (we only accept glued parts if the structure is a demo or trial
version). The final aim will be to design a three-dimensional model of
our structure and manufacture it. The future structure will be made
of metal and we will put a specific attention to make it compact and
62 CHAPTER 4. LIDAR PLATFORM V1

to distribute the mass as close as possible to the axis of rotation of the

servos. Moreover, re-designing the structure we will try other solutions
and a different allocation of the motors.

• + Faster scanning process;

• + Robust;
• + Customized (suits perfectly);
• - Requires time and money to re-adjust;
• - high level of complexity related to the manufacturing process.
Chapter 5

LiDAR platform v2

This is, probably, the most important chapter of this work. In fact, it’s
fundamental to show that we understood problems and issues emerged in
the first version. And furthermore, we have been able to propose a possible
solution that suits our requirements and allows to reach the expected qual-
ity standard. The decision that has been carried out consists in the radical
change from servo motors to steppers. This solution brings as consequence
the need to re-write the code that control pan and tilt movements, adapting
it to the new technology that has been implemented. Beside that, the other
parts will substantially remain the same.
Just to summarize, the big change is related to hardware. This, involves big
changes in the software part related to motion control. Other small software
implementations will be carried on in this phase, but the impact on the final
products is really limited or not substantially relevant. Anyway, consider-
ing big changes, it’s mandatory to conceive a new set of experiments, able
to reveal the reliability and the performances of the new motion system in
particular.
In the second version of our LiDAR, we adopted stepper motors. On pa-
per, this kind of motors can execute very small rotations. If we take for
example, a stepper motor NEMA 171 it can deliver a minimum rotation of
1.8◦ degrees. If a stepper motor is coupled with a micro-stepper driver is
possible to obtain fractions of this angle (typically 1/2, 1/4 , 1/8, 1/32). Since
the micro-stepper represents an open loop control there’s the need to check
and verify the mechanical accuracy of the motors.

Experimental Structure As consequence of what said before, we are

really interested in the validation of the new pan & tilt movement system.
To achieve this, we need to set up a mechanical experiment able to testify the
accuracy of the stepper’s movements. Which are the main characteristics
1
NEMA 17 stepper motor is a stepper motor type specified by NEMA. NEMA 17
steppers have a 1.7 x 1.7 inch faceplate.

63
64 CHAPTER 5. LIDAR PLATFORM V2

Figure 5.1: Experiment step by step

and features for this experiment?

• able to measure precisely;

• be accurate;

• reliable;

• give a repeatable result.

For the experiment we want to follow a precise and well-organized mind set
(See Fig. 5.1). Thanks to that, it will be possible to reduce the failure
rate of the experiment to values close to zero. And at the same time, by
controlling the variables that can affect the result we can get more accurate
and reliable results.
Starting from the problem and setting the goals of our experiment we can
proceed defining the set up. With the word set up we mean the total equip-
ment that we need and the whole procedure that we necessarily need to
follow to get a successful experiment. At this stage, it’s at the same time
important to take into account external factors which can affect the results,
5.1. STEPPER MOTORS 65

such as temperature, presence of dust, humidity, etc. If one of the things we

have just mentioned (and others) will influence in some ways our activity, we
must consider how and when mitigate their effect. Next step, it’s essential
to check the functionalities of the equipment available, calibrate the instru-
ments and correct some biases due to non-zero errors or non-linearities.
When this stage has handed, it’s necessary to write a list containing the
phases that, step by step, constitute our experiment. Strictly following the
process guarantees that the measure will be reliable and repeatable. If then
the result satisfies what was set as the goal of the experiment it’s possible
to say that the process has ended, while if the data elaboration gives un-
expected results it’s necessary to reiterate and try a better set up for the
experiment.

5.1 Stepper Motors

Stepper Motors are used in a wide variety of devices ranging from 3D print-
ers and CNC machines to DVD drives, heating ducts and even analogue
clocks. Stepper motors are DC motors that rotate in precise increments or
”steps”. They are very useful when you need to position something very
accurately. They are used in 3D printers to position the printhead correctly
and in CNC machines where their precision is used to position the cutting
head.
Unlike DC motors, stepper motors are controlled by applying pulses of DC
electricity to their internal coils. Each pulse advances the motor by one step
or by a fraction of a step, the latter is known as ”micro-stepping” and will be
explained shortly. Some people confuse stepper motors with servo motors,
but they are actually two different things. A servo motor is unique in that
its motor shaft can be moved to a precise angle, most servos only rotate
180 or 270 degrees. A servo motor is ”aware” of its position (as seen in the
previous chapter) and can be moved to a specific angle even if an external
force moves the motor shaft.
Steppers, on the other hand, are ”unaware” of their position. They can be
moved to an exact position in reference to where they start stepping but
unlike servos they can be misaligned if their shaft is moved by an external
force. In many applications a servo is first moved to a ”homing” or refer-
ence position before being controlled. Because the move in discrete steps,
a stepper motor is not often used where a smooth continuous rotation is
required, however with the use of gearing and micro-stepping they can ap-
proach a smooth rotation and their ability to be very accurately positioned
often outweighs the roughness of their movement.
Another advantage stepper motors have over DC motors is the ability to
move at very slow speeds without stalling. They also have a high-torque
density, so it means that they can provide a high torque if compared to
66 CHAPTER 5. LIDAR PLATFORM V2

Figure 5.2: Working principle Stepper (a)

Figure 5.3: Working principle Stepper (b)

their size.
Stepper motors have a magnetized geared core that is surrounded by a num-
ber of coils which act as electromagnets. Despite the real number of coils,
electrically, we can draw a simplified scheme that comprehends only two
coils in a stepper motor, divided into a number of small coils. By precisely
controlling the current in the coils, the motor shaft can be made to move
in discrete steps, as illustrated in the following diagrams (Fig. 5.2): In the
first diagram the coil at the top is energized by applying electricity in the
polarity shown. The magnetized shaft is attracted to this coil and then locks
into place. Now look what happens when the electricity is removed from
the top coil and applied to the other coil (Fig. 5.3). The shaft is attracted
to the second coil and locks into place there. The jump between the two
positions is one step (in this illustration a step is 90 degrees, in actual fact
a stepper motor usually steps just a fraction of this. The diagrams are sim-
plified for clarity). We have seen how the motor shaft moves to lock itself
into place in front of an attracting electromagnet, each magnet represents
one step. It is, however, possible to move the motor shaft into positions
between steps. This is known as ”micro-stepping”. In order to understand
5.1. STEPPER MOTORS 67

Figure 5.4: Working principle Stepper (micro stepping)

Figure 5.5: Bipolar Stepper

how micro-stepping works, it’s possible to See Fig. 5.4: In this illustration
the current has been applied to both coils in an equal amount. This causes
the motor shaft to lock into place halfway between the two coils. This would
be known as a ”half step”. The principle can be extended to include quarter
steps, eight steps and even sixteenth steps. This is done by controlling the
ratio of the current applied to both coils to attract the motor shaft to a
position between the coils but closer to one coil than the other. Commonly
this operation is executed by a micro-stepper driver: choose a good quality
driver is quite important to reduce vibrations and for smooth operations.
By using micro-stepping, it is possible to move the shaft of a stepper motor
to a fraction of a degree, allowing for extremely precise positioning.
Actually, it’s possible to divide stepper motors in two categories: bipolar
and unipolar. Since we don’t want to explain everything in detail I’ll just
talk about bipolar stepper (the one we are interested in and the one on
which the experiment is based and conceived). Bipolar stepper motors, See
Fig. 5.5, consist of two coils of wire (electrically, actually split into sev-
eral physical coils) and generally have four connections, two per coil. The
simplified diagrams of stepper operation presented in the previous section
are all bipolar stepper motors. An advantage of bipolar stepper motors is
that they make use of the entire coil winding, so they are more efficient.
68 CHAPTER 5. LIDAR PLATFORM V2

However, they require a more complex controller or driver to operate, as to

reverse direction the polarity of the voltage applied to the coils needs to be
reversed.

5.1.1 Micro Stepper driver and wiring scheme

These chips keep the power that drives the motors separate from the power
that is on the Arduino (in our case, but others micro-controller can be used).
The Arduino can’t provide enough juice to power the stepper motors directly.
This is why it’s necessary to use separate chips to sort of act as valves that
control how the motor spins. Another benefit that stepper driver chips pro-
vide, is that they provide fractional steps, as mentioned before. This helps
smooth out the motion of the stepper motor. Without fractional steps, step-
per motors have a tendency to vibrate or resonate at certain RPMs. To run
a stepper motor, two things are normally required: A controller to create
step and direction signals (at ±5 V normally) and a driver circuit which
can generate the necessary current to drive the motor. In some cases, a
very small stepper may be driven directly from the controller, or the con-
troller and driver circuits may be combined on to one board. The stepper
controller drives 3 wires – traditionally labelled PUL (or pulse), DIR (direc-
tion), GND (voltage reference, sometimes, like in our case, the GND pin is
substituted by the VCC pin and requires a stable 5V voltage reference from
the microcontroller) – which carry motion information to the stepper driver.
Typically, these 3 lines are opto-isolated at the front end of a stepper driver.
The stepper controller is typically a pure digital logic device and requires
relatively little power. The stepper driver (See Fig. 5.6) connects to the
4 thick wires of the stepper motor. It contains the power transistors and
requires a thick power cable to a DC power supply, because all the power
to drive the motors runs through it. The following scheme represents the
wiring diagram of the stepper controller, the stepper driver and the power
source. This sub part of the system is the most important one, since it’s the
one on which we want to focus our attention. Once, we have explained how
the stepper module works we can represent the general set up of the whole
experiment and finally present the results we have gathered. As it’s possible
to see on the front cover of the driver (Fig. 5.7), two tables are reported:
the first one describes the configuration of three switch (SW5, SW6, SW7)
that determines the micro-stepping function. The other table reports the
maximum current that the driver can use to feed the stepper. This value
depends on the specifications and features of the stepper itself. So, for this
choice it should be taken into account the stepper datasheet, and then set
the position of other three switch (SW1, SW2, SW3). Last but not the least
is SW4 responsible to prevent the motor to be damaged by current spikes.
By writing a specific code compatible with the microcontroller, and by set-
ting carefully the micro-stepping configuration on the driver, is possible to
5.1. STEPPER MOTORS 69

Figure 5.6: Micro Stepper Driver (DM320)

Figure 5.7: Stepper Module

70 CHAPTER 5. LIDAR PLATFORM V2

 High contrast & sharp

 Captures full resolution of megapixel cameras
picture in all areas of the screen
 Compact design. Diameter
 Low distortion (Less than 1.0%)
33.5mm (H0514-MP: 44.5mm)
 Locking set screws for focus and
iris

Table 5.1: Camera Features

control and command the stepper motor. In our case we use Arduino as
microcontroller and we flash the board with a specific IDE. Other micro-
controllers can be used, getting a pretty similar result. Right now, it has
been explained how a stepper and a stepper control work, so it’s possible
now to go further presenting the experiment.

5.2 Experiment Set up

Actually, we can say that the experiment’s aim is to evaluate a mechanical
entity, which is the minimum rotation executed by a stepper motor. But
how to do so?
The question can have multiple answers. The approach that has been fol-
lowed in this paper is practically opto-mechanical. In other words, the ex-
periment can reveal, thanks to a high-resolution camera, the orientation of
a carbon fiber pre-preg. By varying the orientation of the fiber sheet is
possible to detect small rotations through the CCD sensor of the camera.
Before starting the measure acquisition process is fair to pose ourselves some
questions: is the system reliable? Is the system able to detect the minimum
angle that we are looking for? So, in other words, which are the limits of
the experimental set up. Since it’s the first time that I personally use this
equipment, I cannot answer these questions in advance. I will dedicate a
specific section related to the topic after the data will be presented.

5.2.1 Camera specifications and features

The CCD sensor that is used in the experiment is a MegaPixel Lens Camera.
It is manufactured by Computar, a Japanese company and presents the
characteristics reported in Tab. 5.1. On the camera’s body are present two
sliding gears: the lower one is responsible of the accommodation process and
the upper one is responsible to set the quantity of light that flows through the
CCD sensor. In other words, the first lens is relative to the accommodation
process that modifies the focus of the camera by varying the distance of the
lenses and the second one is a diaphragm. These two processes, actually,
replicate what the human eye does (See Fig. 5.8).
5.2. EXPERIMENT SET UP 71

Figure 5.8: Camera drawing & features table

5.2.2 Carbon fiber single layer pre-preg

As we have mentioned before, a special importance for the experiment is
related to the carbon fiber pre-preg. Pre-preg is ”pre-impregnated” com-
posite fibers where a thermoset polymer matrix material, such as epoxy, is
already present (See Fig. 5.9). The fibers often take the form of a weave
and the matrix is used to bond them together and to other components
during manufacture. The thermoset matrix is only partially cured to allow
easy handling; this B-Stage material requires cold storage to prevent com-
plete curing. B-Stage pre-preg is always stored in cooled areas since heat
accelerates complete polymerization. Hence, composite structures built of
pre-preg will mostly require an oven or autoclave to cure. The main charac-
teristic of the pre-preg lies in the fiber orientation: as said before fibers are
stretched along the sheet to form a straight line that is than drowned into
resin. Thanks to that the structure is well organized and defined.

5.2.3 Equipment List

For the complete equipment list See Tab. 5.2.
While the Process (See Fig. 5.10) follows these steps:

1. The camera captures the absolute orientation angle of the fiber pre-
preg;
72 CHAPTER 5. LIDAR PLATFORM V2

Figure 5.9: Carbon fiber pre-preg sample

Figure 5.10: Experimental Set up

5.2. EXPERIMENT SET UP 73

Item Function
The aim of the mechanical structure is to
hold the CCD sensor (high-res) camera in
Structure
place, as well as to position the metallic plate
fixed to the stage (following).
A rotary stage is a tool that couples a
stepper motor with a gearing system and
transforms the mechanical rotation of the
Rotary Stage stepper’s shaft to a high precision rotation
of a mechanical stage.
The stage gear ratio is 1:90;
Size of the stage plate: 60mm;
Nema 17
1.8◦ minimum rotation
Maximum
current: 1.3 A
Torque:
0.22 Nm
Code:
Stepper motor & driver
42HAP34BL4
The motor is then coupled and driven by
the DM320, showed and presented above.
If that motor is coupled with a driver that
provides 1/20 micro-stepping, considering
the gear ratio of the stage, it’s possible to get
0.001◦ minimum rotation of the stage.
In a CCD image sensor, pixels are represented
by p-doped metal-oxide-semiconductors (MOS)
capacitors. These capacitors are biased above
the threshold for inversion when image
CCD Camera acquisition begins, allowing the conversion
of incoming photons into electron charges at
the semiconductor-oxide interface; the CCD
is then used to read out these charges.
More details in the dedicated section.
The software interface triggers the
acquisition of the image and saves
Software interface the result in the pc. Then the picture is
processed and the inclination of the fibers is
detected and showed on the screen.
This is the tool that reveals the rotation.
Thanks to its stretched profile it’s possible
Single layer Carbon Fiber prepreg
to detect the rotation of the fiber positioned
firmly on the stage.

Table 5.2: Equipment List

74 CHAPTER 5. LIDAR PLATFORM V2

2. The stepper motor steps by a certain number of steps (this number

is set up precisely by coding the micro-controller) as previously de-
scribed. At the same time, the stage will rotate, and the orientation
of the fiber will change. Now, it’s necessary to take into account the
fact that one step of the stepper motor corresponds to 1:90 degrees
rotation of the platform if the micro-stepping is avoided. Conversely,
the angle should be adjusted dividing by the micro-stepping factor.

Examples:

Stepper minimum step 1.8◦ ;

no micro-stepping /1;
Gear ratio: 1/90;
⇒ rotary stage rotation per step: 0.02◦ ;

Stepper minimum step 1.8◦ ;

micro-stepping 1/2 ;
Gear ratio: 1/90;
⇒ rotary stage rotation per step: 0.01◦ ;

Stepper minimum step 1.8◦ ;

micro-stepping 1/4;
Gear ratio: 1/90;
⇒ rotary stage rotation per step: 0.005◦ .

3. The camera is triggered a second time to capture the new picture of

the fiber orientation;

4. Thanks to simple calculation is possible to obtain the ∆ rotation (be-

tween status 1 and status 2). See Fig. 5.11 to have an intuitive and
visual idea of the whole process.

By knowing the number of steps that the stepper has executed and taking
into account what has been said in point 2), it’s possible to make a compar-
ison between the expected value and the value that has been measured by
the camera with the second picture. One thing is valuable to be underlined:
if we analyze the process we understand that the measure does not depend
on the absolute values gathered by the camera. In fact, we are interested
in the difference between the two values rather than the absolute numbers.
This means that we can tolerate some errors related to non-zero alignment,
since the error will, presumably, affect the two measures in the same way.
This is a very good result, in fact we have as consequence a stable and re-
liable method to measure what we are looking for. The only thing that is
unknown at this point is if the camera will be able to detect a single step
of the whole system. In that case, it won’t be a problem let the motor step
5.3. EXPERIMENTAL DATA 75

by a precise number of steps (i.e. 10, 100, 1000, 10000, etc.) and then
check the amplitude of the average step, by simply dividing the angle that
is measured with the camera by the number of steps. Last but not the least
we need to make some annotations. Talking about the experimental set up,
it’s necessary to say that:

• The fiber should be well positioned and fixed on the rotary platform,
that lies on the rotary stage. In detail, four pieces of thin tape has
been attached on the plate and the Pre-Preg sheet has been firmly
anchored;

• The Pre-preg sheet has been cleaned with some alcohol in order to
remove some pollution and dirt from the surface. This step requires
patience since it’s very easy to damage the carbon fiber layer. So,
if the sheet is damaged the software struggles to recognize the fiber
pattern and at the same time the output measure is not accurate;

• A special light has been used to illuminate the fiber sample, in this
way it’s possible to increase the contrast factor and this helps a lot in
the patter recognition of the fiber;

• As we have mentioned before the camera presents two sliding gears.

It’s important to calibrate the focus and the diaphragm aperture to
increase the resolution and contrast of the picture.

5.3 Experimental Data

In addition of what reported before, we can spend some more words on the
methodology that will be used during the data gathering process. First of
all, a micro-stepping factor will be set, and for each fms , 4 or 5 measurements
will be taken, especially the ones related to 10-100-1000-10000 steps. In that
way, we want to test if the stepper is accurate, but moreover we want to
underline which are the boundaries and the limit of our set up. Only then, it
will be possible to make a diagram that expresses n∆s versus ns itself. Just to
be clear and complete, the process already presented in the previous chapter
will be explained with pictures (only one measurement) and then the other
data will be reported in charts.

5.3.1 Data Report

The first set of measurements is obtained by using a micro-stepping factor
equal to 1/2 . As shown in the following pictures, the camera has been
triggered twice: the first time, as reference, while the second time, after the
rotation, to measure the new angular position of the fiber. In this case the
76 CHAPTER 5. LIDAR PLATFORM V2

Figure 5.11: Computational Process

5.3. EXPERIMENTAL DATA 77

Figure 5.12: Measurement Sample

number of step that has been chosen is 5485. So:

M easure1 = 35.1457 [o ];
M easure2 = 88.1025 [o ];
Result = 52.9568 [o ] ∆(M easure2 − M easure1 );

Right now, the next step is to calculate the average amplitude of one step.
The operation in pretty easy, it’s just necessary to divide ∆ by the number
of steps ns .

52.9568 [o ]
⇒ = 0.00966 [o /step]
5485 [step]
With simple math is possible to calculate the theoretical stepping angle and
the relative error:
1.8 12
= 0.01
90
(0.01 − 0.00966)
× 100 = 3.431%
0.01
As it’s possible to see, the data are quite reliable (See Tab. 5.3). That means
that the experimental set up is able to detect what we wanted to measure.
As already mentioned, in the following data analysis we will underline the
boundaries of the system. By the way, the data prove that the stepper is a
good solution to implement and integrate to the LiDAR project, since it’s
possible to obtain small and accurate rotations. Before showing the second
set of data measurements, it’s possible to show and present some diagrams.
In this way it will be possible to have an immediate idea of the data set.
In the following diagrams (Fig. 5.13 and Fig. 5.14) will be showed the
78 CHAPTER 5. LIDAR PLATFORM V2

Nstep M1 M2 ∆(M2 − M1 ) Avgstep Error (%)

1 14.1264 13.9957 - 0.1307 0.1307 1207 %
10 13.4796 14.1264 0.6468 0.06468 546.8 %
100 12.8213 13.7613 0.94 0.0094 6%
125 67.5169 69.0173 1.5004 0.012 20 %
250 65.0771 67.5169 2.4398 0.0097592 2.408 %
500 59.9773 65.0771 5.0998 0.0101996 1.996 %
1000 10.8079 21.7317 10.9238 0.0109238 9.238 %
5485 35.1457 88.1025 52.9568 0.00966 3.431 %
10000 2.09782 101.1254 99.02758 0.009903 0.9724 %
18000 2.81901 182.29837 179.47936 0.009971 0.29 %

Table 5.3: Data Set 1, fms = 1/2

Figure 5.13: ∆/Ns vs Ns (1/2)

correlation between the average micro-step and the total number of steps.
In this way it’s easy to see that the asymptotic value is quite close to the
expected value (theoretical step, as computed before). The second diagram
(Fig. 5.14) shows a zoomed area close to the origin of the axis. The
system turns out to be not accurate considering small rotations and small
angles. Especially for a number of step that is smaller than 100. That
means that when the rotation is smaller that 1◦ the experimental set up
struggles to detect the orientation of the fiber. The next step is to present
the second data set, to see if the data are confirmed or not and then propose
a motivation of the inaccuracy of the measurements. As said, the second
set will be acquired setting the micro-stepping factor to 1/4 (Tab. 5.4). And
the relative diagram is the one reported in Fig. 5.15. In this case too,
the experiment shows an inadequacy to detect very small rotations, but
asymptotically it reaches the expected value of 0.005◦ .
5.3. EXPERIMENTAL DATA 79

Figure 5.14: ∆/Ns vs Ns (1/2) - Detail

Nstep M1 M2 ∆(M2 − M1 ) Avgstep Error (%)

10 23,0919 23,0354 -0,0565 -0,005650 43,5000%
100 22,9508 23,0919 0,1411 0,001411 85,8900%
200 31,5734 32,5250 0,9516 0,004758 4,8400%
1000 2,2984 6,1492 3,8508 0,003851 22,9832%
9000 59,9773 104,8369 44,8596 0,004984 0,3120%
10000 6,1492 56,9976 50,8484 0,005085 1,6968%

Table 5.4: Data Set 2, fms = 1/4

Figure 5.15: ∆/Ns vs Ns (1/4)

80 CHAPTER 5. LIDAR PLATFORM V2

Figure 5.16: Carbon Fiber Sample Detail

5.3.2 Conclusions

• The data show that the system is reliable and helps us to check what
we wanted to check. In this way is possible to test the accuracy of a
stepper motor. This was the initial goal;

• It’s possible to see that the values are quite similar to the ones we
expected to find, at least for a sufficient big number of steps. It’s
possible to inflict that, when the rotation of the stage is smaller than
1◦ , the relative error is huge. A possible answer could be related to
the average operation. In fact, with small angles the error is split on a
small number of steps, while with a bigger angle the error, related to
the detection of the carbon fiber orientation, is split on a big number
of steps. This bring us to think that the algorithm that is used to
identify the carbon fiber pattern could be revised and it can become
more robust and reliable. The algorithm has proved to be sufficiently
efficient with quite wide rotations, not the same with small or super
small ones;

• It’s very important to set the camera with the right diaphragm aper-
ture and focus, and try to increase as much as possible the luminance
5.4. NEW STRUCTURE 81

contrast:
(Ls − Lb )
Cw =
Lb
Where Cw indicates the luminance contrast, Ls the luminance of the
target object and Lb the luminance of the background;
• It’s important too to pay attention to the positioning of the carbon
fiber on the metallic plate. Some vibrations, or accidental collision
(with other structural parts) can modify its position and give wrong
results;
• At the same time is relevant the quality of the carbon fiber that is
used. Generally, when the carbon fiber is flexible and malleable it’s
more difficult to recognize the pattern. That’s why is recommended to
partially activate the resin through heat, UV light or other equivalent
methods;
• The adoption of the stepper technology has proved to be the right
solution to improve the density of the LiDAR point cloud. By coupling
ad-hoc electrical driving systems and mechanical system is possible
to obtain densities close to 1Million points, and at the same time
guarantee a high level of accuracy.

5.4 New Structure

Another terrific improvement on LiDAR platform v2 is related to the struc-
ture. Actually, there’s no much to say: the new structure is solid and more
firm. It allows a better positioning of the sensor as well a better general
structure stability. The structure has been firstly drawn in SolidWorsTM and
then manufactured in Steel. In order to improve the overall system perfor-
mances and reduce the minimum stepper angle we used a mechanical so-
lution. On the pan stepper’s shaft is mounted a gear system. The small
gear is linked to the pan motor, while the big one is linked to a secondary
shaft on which the a small platform is mounted. This platform has been
conceived to be the housing of the tilt motor as well as the sensor’s housing.
Reducing the speed allows to reduce contextually the minimum angle that
the pan stepper delivers. The two gears have a ratio 1:5 and they’re mu-
tually connected through a belt, specifically designed. Here we have some
pictures: See Fig. 5.17a and Fig. 5.17b.

5.5 Results and Scan

In this section the final version of the LiDAR will be presented, showing the
relative pictures (See Fig. 5.18a and Fig. 5.18b). Moreover, the results of
further scan will be showed and then commented.
82 CHAPTER 5. LIDAR PLATFORM V2

(a) Mechanical Assembly (b) Mechanical Assembly Exploded

Figure 5.17: Mechanical Assembly, Detail

(a) LiDAR v2, Detail

(b) LiDAR v2, Detail

Figure 5.18: LiDAR v2

5.5. RESULTS AND SCAN 83

(a) Real Picture

(b) Point Cloud LiDAR v2

Figure 5.19: Table and Boxes Scan, v2

84 CHAPTER 5. LIDAR PLATFORM V2

(a) Real Picture

(b) Point Cloud LiDAR v2

Figure 5.20: Living Room Scan, v2

5.5. RESULTS AND SCAN 85

Figure 5.21: Human Body Scan

86 CHAPTER 5. LIDAR PLATFORM V2
Chapter 6

Conclusions

This last chapter is fundamentally a recap of what has been treated and
presented in this work. The main goal of the project was to develop a
versatile LiDAR platform able to scan the surrounding environment with a
decent accuracy and quality of the point cloud. We repeat another time that
quality of the point cloud, actually means density: the more dense a point
cloud is, the more accurate the final representation is. Anyway, the main
goal was to meet these requirements, dealing with a constraint on costs. The
final platform, indeed, should have cost roughly 250$. With the following
list the whole process will be recapped and some conclusions or notes will
be added:

1. We started from the basic knowledge, from scratch, studying the single
components and parts of the project, learning the theory and trying
to figuring out how to solve a complex problem as the one we had in
input;

2. Finally, the LiDAR v1 has been designed first and then manufactured:
the development of this platform has been challenging and required us
to put together a vast knowledge and deploy multidisciplinary skills;
basically related to servo motor control, Arduino servo management
and integration, the LiDAR sensor and other various equipment. But
it’s not finished, indeed, the project required us to develop from scratch
a brand new visualization environment which allows us to show the real
time point cloud that represents the virtualization of the surrounding
reality. This environment has been written on a Java platform called
Processing v3 (virtual image processing system);

3. Unluckily, LiDAR v1’s performances were not as good as expected. So,

after the testing phase, we focused our attention on which were the
limits of the first version and which were the updates or adjustments
thanks to which LiDAR’s performances could have improved. Only

87
88 CHAPTER 6. CONCLUSIONS

then, the most efficient and profitable solutions has been adopted in
version 2;

4. Then, LiDAR v2 has been released. This second version required an

impressive effort, in particular due to the new pan & tilt motor module
(brand new technology) and the design of the new platform’ structure.
As said, moving from servos to steppers has been a huge leap. So we
had to face lots of unknowns and doubts. In order to understand if the
stepper actually suited our application, we conceived a test carried on
in the lab, aimed at testing these components, in terms of reliability,
accuracy and performance. The test gave us the answers we were
looking for. Beside that, as hinted, a new structure has been released.
A final test phase then proved huge improvements in real word scenario
representation.

5. Talking about the LiDAR prototype and its possible applications, it’s
quite easy to say 3D virtual scan and point cloud representation of
interiors (what we did in the test phase). But then, with further
improvements and updates that machine could be used as a low cost
platform applicable to:

– Urban planning and modelling (new infrastructures planning in

critical existing contexts);
– 3D Building modelling;
– Infrastructure monitoring and control (status check of power lines,
railways, ...). This last application is strictly related to my major
and my academic background, so it takes on a truly interesting
meaning.
90 CHAPTER 6. CONCLUSIONS
Bibliography

[1] Robert K. Wysocki, Effective Project Management. Traditional, Agile,

Extreme. Fifth edition(2009). ISBN 978-0-470-42367-7.

[2] LIDAR—Light Detection and Ranging—is a remote sensing method used

to examine the surface of the Earth”. NOAA. Archived from the original
on June 4, 2013. Retrieved June 4, 2013.

[3] J. Young, Lidar for Dummies, (2011), Wiley Publishing Inc., ISBN: 978-
0-470-94225-3.

[4] Cracknell, Arthur P.; Hayes, Ladson (2007) [1991]. Introduction to Re-
mote Sensing (2 ed.). London: Taylor and Francis. ISBN 0-8493-9255-1.
OCLC 70765252.

[5] M. Doneus, I. Miholjek, G. Mandlburger, N. Doneus, G. Verhoeven, Ch.

Briese, M. Pregesbauer, ”Airborne laser bathymetry for documentation
of submerged archaeological sites in shallow water”. ISPRS – Interna-
tional Archives of the Photogrammetry, Remote Sensing and Spatial
Information Sciences. Bibcode:2015ISPArXL55...99D.

[6] Chiu, Cheng-Lung; Fei, Li-Yuan; Liu, Jin-King; Wu, Ming-Chee. ”Na-
tional Airborne Lidar Mapping and Examples for applications in deep
seated landslides in Taiwan”. Geoscience and Remote Sensing Sympo-
sium (IGARSS), 2015 IEEE International. ISSN 2153-7003.

[7] G. Vosselman,H. Maas, (2012). Airborne and terrestrial laser scanning.

Whittles Publishing. ISBN 978-1-904445-87-6.

[8] S. Lee, J. Joon Im, B. Lee, Leonessa, K. Alexander. ”A real time grid-
map generation and object classification for ground based 3D lidar data
using image analysis techniques”. Image Processing (ICIP), 2010 17th
IEEE International Conference on image processing. ISSN 1522-4880.

[9] A. Jadhav, D. Gambhir. ”Use of LiDAR in Extraction of 3D City Models

in Urban Planning”. September 1, 2009. Geospatial World Journal.

91
92 BIBLIOGRAPHY

[10] J. A. Jakubiec, C. F. Reinhart. ”A method for predicting city-wide

electricity gains from photovoltaic panels based on LiDAR and GIS data
combined with hourly Daysim simulations”. 12 May 2013, ELSEVIER
Solar Energy 93 (2013) 127–143.

[11] D. Lv, X. Ying, Y. Cui, J. Song, K. Qian, M. Li. ”Research on the

Technology of LIDAR Data Processing”. The 28th Research Institute of
China Electronics Technology Group Corporation Nanjing, China.

[12] NTSB 2016 Report:”Highway Deaths Lead National Increase in Trans-

portation Fatalities”. 11.27.2017. https://www.ntsb.gov/news/press-
releases/Pages/PR20171121.aspx

[13] The picture is available here. https://www.continental-

automotive.com/Landing-Pages/CAD/Automated-
Driving/GlobalHighlights/3D-Flash-Lidar (Continental Website).

[14] Extract from here. https://www.cosworth.com/news/electronic-

news/demonstrating-the-car-of-the-future-with-lidar-sensor-technology

[15] Lidar Lite v3 Operation Manual and Technical Specifications. GARMIN

[16] The I2C-Bus Specification Version 2.1. Philips Semiconductors. Jan-

uary 2000.

[17] ttps://www.sciencebuddies.org/science-fair-
projects/references/introduction-to-servo-motors

[18] https://processing.org/

[19] Ira Greenberg (31 December 2007). Processing: Creative Coding and
Computational Art. Apress. pp. 151–. ISBN 978-1-4302-0310-0.

[20] Jeanine Meyer (15 June 2018). Programming 101: The How and Why
of Programming Revealed Using the Processing Programming Language.
Apress. pp. 121–. ISBN 978-1-4842-3697-0.

[21] Ira Greenberg (25 March 2010). The Essential Guide to Processing for
Flash Developers. Apress. pp. 412–. ISBN 978-1-4302-1980-4.

[22] Casey Reas and Ben Fry. Processing: A Programming Handbook for
Visual Designers, Second Edition. December 2014, The MIT Press. 720
pages.