Technical Paper of Sensors & Transducer

Assign. 1, No. 1, Sem1, 2014

Ultrasonic Sensors for Obstacle or Object Detection and Avoidance

1

Ling Nai Ho1

Department of Control and Mechatronics Engineering, Faculty of Electrical
Engineering, Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor.
[email protected]
Abstract

This paper presents the review of nine most recent research technical papers from various
Journal papers in development of ultrasonic sensors for obstacles or object detection and
avoidance. Various techniques and modifications have been proposed and tested on
ultrasonic sensors to eliminate errors and improve performance of sensors in various fields
ranging from robotics automation to aviation field with lowest cost available.
Keywords: Ultrasonic sensor, high directivity, parametric array, obstacles detection, distance
measurement, beam overlapping.

1. Introduction
Nowadays, the technology of distance measurement and object detection as well as
collision avoidance are in high demand in various fields ranging from medical industries to
automation and even aerospace industries. Normally, ultrasonic sensors works with simple
and low-cost principle either with an on or off digital output for detecting objects movement
or with an analog output which varies proportionally to the sensor to target separation
distance (Satish et al., 2008). Numerous ways have been developed throughout decades to
improve and diversify the usage of ultrasonic sensors in diverse industries such as material
handling, mobile equipment, fetus and organ scanning, sonochemistry, aircraft ice detection,
flaw detection, fill level measurement, UAV navigation, automatic parking and etc.
However, in order to achieve the expanded functions mentioned above, researches and
experiments are carried out on ultrasonic sensors together with combined operations of other
sensors such as laser, optical sensors, proximity, infrared and etc depending on the field
involved. (Huihai C., Jinze S., Daxue L, 2013). This paper only focuses on the field of
object detection and obstacles avoidance and thus, discussing mainly on the problems
arised from technical resolutions of ultrasonic sensors as well as different designs as
solutions.

2. Theory
Ultrasonic sensors are defined as electronic devices that emit an aco ustic wave
beyond the upper range of human hearing -- called the audible range, between 20 hertz
and 20 kilohertz. To generate the ultrasonic wave, ultrasonic sensors use a vibrating
device known as a transducer to emit ultrasonic pulses that travel in a cone-shaped
beam. The range of an ultrasonic sensor is determined by the frequency of vibration of
the transducer. As the frequency increases, the sound waves transmit for progressively
shorter distances and transmit for longer distances as frequency decreases. Thus, longrange ultrasonic sensors work best at lower frequencies, and short -range ultrasonic
sensors work best at higher frequencies.

Technical Paper of Sensors & Transducer

Assign. 1, No. 1, Sem1, 2014

Ultrasonic sensors come in a variety of configurations and typically use one or more
transducers, depending on the application. Spacing between multiple transducers is an
essential characteristic to consider since unwanted interference may occur if the transducers
are spaced too closely together. In the current application, ultrasonic signals are used for
obstruction detection by producing a burst, and detecting its reflections after they hit the
obstruction. Basically this is the technique used by bats to travel in the nights (Satish et al.,
2008).
2.1 Principle
For measuring the distance, time of flight for reflected ultrasonic signals is calculated.
Ultrasonic signals have good directionality, and they are easily reflected by solid objects. By
finding the time taken for ultrasonic signals to reflect, the distance traveled by sound can be
easily calculated, as velocity of sound is a known quantity. This is known as pulse echo
technique, which is used here for distance measurement.

3. Literature Review
The literature review will be about different technologies and methods used in solving the
measurement errors and enhancing the directivity as well as robustness of ultrasonic sensors
in obstacles detection and collision avoidance.
3.1 Parametric Array Method

Generally, ultrasonic waves have better directivity than low-frequency waves which
enables rapid attenuation. Since most of the available commercial polaroid-type sensors
which work on only one working frequency makes it rather impossible to improve its
directivity and working distance. Therefore, directivity related to the spatial resolution as well
as the working distance should be placed at their analytic center. Parametric array method
which uses the frequency difference between two ultrasonic waves is used to generate a
highly directional low-frequency wave with a small aperture (Majchrzak et al., 2009). In
order to develop this sensor with higher directivity while maintaining working distance,
Jongkyu et al. (2009) proposed a new design of transducer which generates two simultaneous
longitudinal modes efficiently that applies the parametric acoustic array.
In their research, a modified version of Gallego-Juarezs ste ed- ate tra sducer ca ab e
of ge erati g dua -fre ue c waves is deve o ed a d tested with the ara etric arra
ethod
to obtai better directivit a d accurac of
co ared to available commercial
roducts with 20

Figure 1.1: Experimental setup for measurement of directivity and frequency response.
Source: (Jongkyu et al. 2009)

Technical Paper of Sensors & Transducer

Assign. 1, No. 1, Sem1, 2014

3.2 Double Vector Field Histogram Method (DVFH)

In application of real-time obstacle avoidance for an omnidirectional mobile robot, a ring
of 32 ultrasonic sensors have been used together with a modified version of Vector Field
Histogram (VFH) method, called Double-VFH (DVFH) which calculates simultaneously two
vector field histograms centered at two points along the longitudinal axis of the robot. The
VFH method compensates for the inaccuracies of the ultrasonic sensors and thus providing
sufficiently detailed spatial representation of the robot's environment for travel among
densely cluttered obstacles without compromising the system's real-time performance.
However, the oscillation tendency when applying on a large, rectangular-footprint platform
robot (Omnimate) was observed in high obstacle density environments which cause
instability to the system during each sampling interval.

Figure 1.2: Large Omnidirectional robot (Omnimate) and snapshot of a momentary polar
histogram for a sample obstacle course with three obstacles A, B, and C.
Source: (Hong Y. et al. 2008)
Hong Y., Johann B. and David W. (2008) proposed the DVFH method which introduces
the modification of the original VFH method for the 3DOF, elongated rectangular-shaped
Omnimate which works based on two trucks, which are kinematically simple differential
drive vehicles. The DVFH algorithm treats these two trucks identically, that is, either one
may lead or follow. Thus, eliminating the oscillation tendency in VFH method.
3.3 Overlapped Ultrasonic Sensor Ring

An ultrasonic sensor can measure distance of an obstacle within its conic beam but
presence of uncertainty arc or measurement errors may cause performance degradation. The
lack of information about measuring errors complicates the evaluation of the reliability of the
final results for obstacle avoidance (Jarosaw M. et al., 2009).
To overcome the measurement errors and uncertainties, a set of ultrasonic sensors of the
same type are arranged along a circle of non-zero radius at regular spacing with their beams
overlapped (Sungbok and Hyunbin K., 2012). By comparing the obstacle distances of three
adjacent ultrasonic sensors, the positional uncertainty of the center ultrasonic sensor can be
accessed and represented in terms of uncertainty sub-arcs.
In addition, Nils G. et al. (2012) proposed that two ultrasonic sensors for one half of the
same angle which results in double redundancy and resolution are always recommended in
obstacles detection and avoidance in Autonomous unmanned-aerial vehicle. That is, 12
senosrs are needed for a 360 degree circle. Hence, both Nils and Sungbok agreed that
overlapped ultrasonic beams are essential in reducing positional uncertainty and errors.
Apart from that, Adrian B. et al. (2010) also uses the same principle of overlapping
ultrasonic beams together with Infrared sensors to accuratel y detect objects at different

Technical Paper of Sensors & Transducer

Assign. 1, No. 1, Sem1, 2014

positions and distances in their research of ultrasonic and IR safety device for mobility
scooter collision avoidance.

Figure 1.3: Beam overlapped using multiple ultrasonic sensors for better positional accuracy
and higher resolution and redundancy in three different projects respectively.
Source: (Sungbok and Hyunbin K., 2012, Nils G. et al. 2012, Adrian B. et al., 2010)

4. Results and discussions

From the research done by Jongkyu et al. (2009), a new type of high directio a
u traso ic ra ge se or is fabricated a d is ab e to ge erate a ara etric acoustic arra with
which is far better tha the co
ercia se sors with 20
The sound
pressure level of the primary wave was more than 120 dB and the difference frequency wave
was 83 dB which is much better than other conventional ultrasonic range sensors.

Figure 1.4: Directional pattern of two primary waves and difference frequency wave.
Source: (Jongkyu et al., 2009)
On the other hand, research done by Hong Y. et al. (2008) using the DVFH method in
Omnidirectional robot is able to traverse even densely cluttered obstacle coursed at a
maximum speed of 0.4 m/sec, although the average speed was only 0.2 - 0.3 m/sec because
the robot slowed down whenever it came close to obstacles. This is due to the real-time
processing delay arised from the two calculations on DVFH algorithm to determine the
motion mode for each trucks either leader or follower. Higher maximum speeds were possible
in less densely cluttered environments [3]. Though accurate and precise, this method is costly
and slower since it requires a ring of 32 Polaroid ultrasonic sensors and a fast processing

Technical Paper of Sensors & Transducer

Assign. 1, No. 1, Sem1, 2014

electronic microprocessor to carefully compute and process the vast amount of data collected
which may prompt to errors in fast robot applications.

On the contrary, the overlapped ultrasonic sensor ring or overlapped beam proposed by
Sungbok and Hyunbin K. (2012) is of higher preference and popular in obstacles detection
and avoidance application. This method is able to produce better resolution and minimal
positional uncertainty despite its cheaper cost and simpler working principle. Moreover,
Sungbok and Hyunbin K. managed to achieve both simple and complex obstacle detection
using an overlapped beam where the uncertainty arc of each sensor can be divided into three
smaller sub-arcs and can be used for detecting from single up to omnidirectional obstacle
detection which is a better and easier approach than that of Omnidirectional robot proposed
by Hong Y. et al. (2009).
To be clear, the positional uncertainty is determined through the combination of ultrasonic
sensors detecting an obstacle for single object detection while multiple obstacle detection is
based on inequality relationship among three consecutive obstacle distances. Lastly, for
omnidirectional obstacle, its positional uncertainty is obtained by overlapping all the
uncertainty arcs obtained for every three consecutive ultrasonic sensors. The summarized
results of overlapped ultrasonic ring experiment are as shown below:

Figure 1.5: Single object detection using three adjacent ultrasonic overlapped beams.
Source: (Sungbok and Hyunbin K., 2012)

Table 1: Decision rules for the positional uncertainty in single object detection
Detailed results presentation of multiple and omnidirectional object detection can be referred
in journal written by Sungbok and Hyunbin K. which shows the procedure of getting the
positional uncertainty and their respective decision rules applied [4].

Technical Paper of Sensors & Transducer

Assign. 1, No. 1, Sem1, 2014

5. Conclusion
Ultrasonic sensors have becoming more and more crucial in industrial applications and
even in our daily lives automatic parking solutions. Therefore, researchers have
proposed and eventually commercialized lots of various ultrasonic sensors just to meet
the market demand. However, from the development of technology and concept
introduced in application of obstacles detection and avoidance such as the parametric
array method, Vector Field Histogram method, Double Field Hi stogram method, AVR
microcontroller ultrasonic sensor, IR and Sonar based senor and overlapped ultrasonic
ring sensor ring or beams.
Nevertheless, overlapped ultrasonic ring senor or beams stand out to be the most
preferable and recommendable method among the researchers nowadays due to its
unexpectedly simple and low-cost operation. Therefore, an overlapped low directivity
ultrasonic senor ring can be a promising solution for low cost yet high performance
obstacle detection and avoidance robot applications.

6. References
[1] Cui, H., Song, J., & Liu, D. (2013, January). Ultrasonic Array Based Obstacle Detection in
Automatic Parking. In Proceedings of 2013 Chinese Intelligent Automation Conference, pp. 129-140.
[2] Park, J., Je, Y., Lee, H., & Moon, W. (2010). Design of an ultrasonic sensor for measuring distance
and detecting obstacles. Ultrasonics, 50(3), 340-346.
[3] Yang, H., Borenstein, J., & Wehe, D. (2000, September). Sonar-based obstacle avoidance for a
large, non-point, omni-directional mobile robot. In Omni-directional Mobile Robot, Spectrum 2008
International Conference on Nuclear and Hazardous Waste Management, pp. 24-28.
[4] Kim, S., & Kim, H. (2012, October). Simple and complex obstacle detection using an overlapped
ultrasonic sensor ring. In Control, Automation and Systems (ICCAS), 2012 12th International
Conference on, pp. 2148-2152.
[5] Majchrzak, J., Michalski, M., & Wiczynski, G. (2009). Distance estimation with a long-range
ultrasonic sensor system. Sensors Journal, IEEE, 9(7), 767-773.
[6] Gao, H., & Rose, J. L. (2009). Ice detection and classification on an aircraft wing with ultrasonic
shear horizontal guided waves. Ultrasonics, Ferroelectrics and Frequency Control, IEEE Transactions
on, 56(2), 334-344.
[7] Bingham, A., Hadoux, X., & Kumar, D. K. (2014, May). Implementation of a safety system using ir
and ultrasonic devices for mobility scooter obstacle collision avoidance. In Biosignals and Biorobotics
Conference (2014): Biosignals and Robotics for Better and Safer Living (BRC), 5th ISSNIP-IEEE, pp.
1-5.
[8] Gageik, N., Mller, T., & Montenegro, S. (2012). Obstacle Detection And Collision Avoidance
Using Ultrasonic Distance Sensors For An Autonomous Quadrocopter. University Of Wurzburg,
Aerospace Information Technology (Germany) Wurzburg September.
[9] Pandey, S., Mishra, D., Srivastava, A., Srivastava, A., & Shukla, R. K. (2008). Ultrasonic
Obstruction Detection and Distance Measurement Using AVR Micro Controller. Sensors &
Transducers journal, 95(8), 49-57.