See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/315576320

Smart, low power, wearable multi-sensor data acquisition system for

environmental monitoring

Article · March 2017

CITATION READS

1 264

4 authors, including:

Serbanescu Matei Ovidiu Emanuel E. Hutanu

Polytechnic University of Bucharest Polytechnic University of Bucharest
6 PUBLICATIONS   13 CITATIONS    5 PUBLICATIONS   39 CITATIONS   

SEE PROFILE SEE PROFILE

Cristian Ravariu
Polytechnic University of Bucharest
211 PUBLICATIONS   474 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

A new Master - Nano-Bio-Engineering and Environment Management - wait candidates from 01. Sept. 2019 at Univ. "Politehnica" Bucharest, Fac. Electronicsronica View
project

Microgrid Project with deadline 20 Jan 2020 View project

All content following this page was uploaded by Serbanescu Matei on 24 March 2017.

The user has requested enhancement of the downloaded file.

 th
THE 10 INTERNATIONAL SYMPOSIUM ON ADVANCED TOPICS IN ELECTRICAL ENGINEERING
March 23-25, 2017
Bucharest, Romania

Smart, low power, wearable multi-sensor data

acquisition system for environmental monitoring
M. SERBANESCU1, V. M. PLACINTA1,2, O. E. HUTANU2,3 , C. RAVARIU1
1
Faculty of Electronics, Telecommunications and Information Technology,
Polytechnic University of Bucharest,
Splaiul Independentei 313, RO-060042, Bucharest, Romania
2
Department of Elementary Particle Physics,
Horia Hulubei National Institute for R&D in Physics and Nuclear Engineering,
Reactorului 30, RO-077125, P.O.B. MG-6, Bucharest-Magurele, Romania
3
Faculty of Electronics, Telecommunications & Information Technology,
“Gheorghe Asachi” Technical University of Iasi,
Bd. Carol I 11, RO-700506, Iasi, Romania
[email protected], [email protected]

Abstract - The wireless sensor network is becoming a such as system integration and low-power sensors interface
technology of huge interest for a wide variety of applications. circuitry design. That is the main reason driving the research
Despite the technological developments of smart sensing, issues and development labor conducted upon this [2].
related to system integration, sensor miniaturization and This work presents the prototype of a smart, low–
especially low-power sensors interface circuitry design, still
power, wearable system that could be used for environmental
represent major technical challenges. This paper describes the
prototype of a smart, low power, wearable system that can be monitoring and geographic orientation and has its principal
used for environmental monitoring. The wearable system uses an concern the CO2 assessment in different surroundings.
array of sensors connected to a central processing unit with LabVIEW plays here a central role, facilitating the data
firmware for continuously monitoring. The data acquisition acquisition and the Bluetooth communication with the
process has been performed via Bluetooth with a Graphical User hardware.
Interface (GUI) designed within LabVIEW. This system will be There is a variety of methods that can characterize
also interfaced with an Android device as an alternative for ease and quantify the global increase in CO2 mixing ratios in the
of access and portability. This work is meant to prove basic atmosphere and their links to anthropogenic emissions and
concepts about smart sensing in the context of wearable systems
terrestrial and oceanic sinks. However, quantifying, mapping,
along with Bluetooth communication, data acquisition as well as
mobile computers. and monitoring CO2 emissions are still a challenge [3].
The related work and current state of the art refers to the
Keywords – Smart, low-power, data acquisition, sensors, uRADMonitor global environmental monitoring network. It
environment. has been designed as a worldwide network of automated
monitors, a first line detection and warning system against the
I. INTRODUCTION & STATE OF THE ART harmful chemical and physical factors around us [4].
Wearable electronics and especially Wireless Sensor This paper is organized as follows. In Section II, we
Network field have become more ubiquitous nowadays, present a brief description of the system along with its main
carrying more sensors, due to the advances in miniaturization features. Section III describes the hardware architecture
and manufacturing. At this moment, technology is at the point consisting of three blocks: sensors block, communication
where these smart devices can embed various kinds of sensors block and power supply block. The experimental procedures
and continuously monitor the environment [1]. and data sets are being discussed in Section IV and the
Despite the potential advantages of monitoring conclusion is drawn in Section V as well as possible future
system relying on sensors like the one to be described, there work.
are significant limitations ahead before such a system can be
used on a large scale. Those involve technological barriers

78-1-5090-5160-1/17/$31.00 ©2017 IEEE

 II. SYSTEM DESCRIPTION & FEATURES
The data acquisition system (DAQ) described in for the CO2 concentration level measurements. The CO 2
this paper can be regarded as an entry level solution for basic measurements are based on non-dispersive infrared method
environmental survey. Besides its portability, this DAQ can (NDIR) [7], which can be self-calibrated and outputs data
be interfaced with almost all mobile computers which embed using ASCII characters via an Universal Asynchronous
a Bluetooth chip in their hardware. Receiver/Transmitter (UART) communication.
The following features are included in the package: For atmospheric pressure measurements the perfect
 Temperature and humidity; candidate was BMP180 sensor [8], which is a dedicated
 Atmospheric pressure and approximate altitude digital pressure sensor. Its pressure range is from 300 hPa to
calculation; 1100 hPa with a medium accuracy of ±1 hPa. Since the
 CO2 concentrations levels; pressure value is relatively inverse linear with the altitude,
 Earth’s magnetic field measurements along with see fig. 2, and based on the mean value of pressure at sea
magnetic North pole indication; level (0 m elevation) the approximate altitude can be
 3-axis acceleration measurements; estimated with the barometric formula [9], shown in equation
 3-axis angular velocity measurements; (1),
 It is powered by a Lithium-Ion battery.

The system design was geared towards low cost, low- p= (1)
power, wearable and a smart solution that can be interfaced
with other electronic systems desired for environmental
monitoring. Its power consumption is around 300 mW in full
transmission mode. where h is the elevation level [m], T0 is the standard value
temperature at sea level, 288.15˚K (15˚C), p0 is the
III. ARCHITECTURE atmospheric pressure at sea level.
The general hardware architecture of this DAQ
system, shown in fig. 1, could be divided into 3 main blocks:
the sensors block, the communication block and the power
supply block.

Fig. 2 Pressure altitude versus elevation level adapted from [9]

The Earth’s magnetic field is measured with the

help of a surface-mount, multi-chip module, HMC5883L
[10]. This sensor embeds a low-field sensing on all 3-axis
Fig. 1 DAQ architecture and is suitable for low cost compassing and magnetometry.
Besides his wide magnetic field range, ±8 Gauss, and with
A. Sensors block the help of mathematics a compass heading can be
The entire block uses 4 sensors to fulfil the determined from the readings of the x and y axis of the
measurements listed in the previous chapter. Earth’s magnetic field [11].
An SHT21 [5] sensor is used to perform precise The last 2 parameters, acceleration and angular
measurements of the environmental temperature and velocity, are obtained using an integrated 6-axis motion
humidity. Its performances are the following: temperature tracking device that combines a 3-axis accelerometer, a 3-
range from -40 °C to +125 °C with a medium accuracy of axis gyroscope and a digital motion processor (DMP). This
±0.5 °C and humidity range from 0 %RH to 100 %RH with a device, MPU-6050 [12], embeds 2 digital communications
medium accuracy of ±2 %RH. This sensor is embedded into protocols, I2C, which can be interfaced with a broad range of
other encapsulated sensor, COZIR-AJ-5000 [6], being digital sensors.
responsible
B. Communication block
The communication between DAQ and mobile
computer has been done via a wireless protocol implemented
using a low cost Bluetooth adapter [13] shown in fig. 3. The
adapter emulates a standard UART port which runs at 115.2
kbps.

Fig. 3 Bluetooth adapter [13]

C. Power supply block

This block contains 2 main components: a Boost
converter and a Lithium-Ion battery charger.
The Boost converter outputs 5 V for an input
voltage range between 0.7 V and 5 V. It has been designed
around MAX756 [14], and can deliver up to 200 mA. This
solution has been chosen because the input voltage from the
Lithium-Ion battery can vary from 3 V when is discharged to
Fig. 5 Software states flowchart
4.2 V when is fully charged.
Apart from that, a mini-USB connector has been
introduced in order to power the entire device and charge it A GUI has been developed within LabVIEW, as
in the same time. Fig. 4 shows the first iteration of the DAQ shown in fig. 6. That allows the user to configure the DAQ
system and the PCB has been designed on two layers. and to read its data every 2 seconds. Data is received as a
well-defined string of ASCII characters, and after some
processing steps, is being displayed on waveform charts. In
background data can be stored, at the user wish, into ASCII
files for later analysis. The entire acquisition, processing,
waveform displaying and storing data stages are governed by
a Finite State Machine Architecture (FSMA) implemented in
the LabVIEW diagram block.

Fig. 4 DAQ assembled and uncased

The entire DAQ system is controlled by a low cost

8-bit microcontroller, ATmega328P [15].This
microcontroller has the main following features: 14 GPIO
(General Purpose I/O) pins, 6 multiplexed analog inputs, 2
KBytes of internal SRAM, 1 KBytes EEPROM, hardware
UART module, hardware SPI module, hardware I2C module,
timers etc.
The software has been written within C
Fig. 6 LabVIEW interface
programming language using the Arduino compiler and its
flowchart structure as shown in fig. 5.
 IV. MEASUREMENTS
The device capability has been evaluated using a
testing strategy that included three scenarios:
 Indoor measurements;
 Outdoor measurements;
 Indoor- outdoor measurements.
Apart from that, a few self tests were conducted in order
to test the readings and functionality of the accelerometer,
the gyroscope and the magnetometer.

A. Indoor measurements
First set of measurements has been completed into
inhabited and unventilated 20 room, where the DAQ was
installed.
In fig. 7 the variation of the 4 most important
parameters like temperature, humidity, atmospheric pressure Fig. 8 Outdoor measurements
and CO2 concentration levels which have been continuously
monitored for about one hour could be seen. CO2 concentration has drop to a constant value
around 400 ppm, and keeps in range with the normal
levels stated by literature [16].

C. Indoor - outdoor measurements

Last set of data has been realized with the purpose
to check the reliability of the DAQ below zero temperatures.
For that, the device was first placed inside at about 27 ˚C
room temperature for about three minutes, and afterwards
was taken outside in fresh cold air at about -6 ˚C. The results
are shown below in fig. 9, hence things are in fully
concordance with the environmental parameters.

Fig. 7 Indoor measurements

As can be seen in the graph, the CO2 levels are

about 1500 ppm, and according with the literature [16] it
is a regular level for this kind of environments which is
also not harmful for human beings.
The atmospheric pressure value is about 764
mmHg, at an approximate elevation level of 90 m above
sea level. This data was taken sometime in November
2016.

B. Outdoor measurements Fig. 9 Indoor - outdoor measurements

The second set of the measurements was done in the
same area but outside of the building and a month later. Apart from these environmental measurements, the
The same strategy has been followed and the results other features concerning the acceleration, angular velocity
could be seen in fig. 8. and the magnetic field have been tested. The magnetic field
Now, it is obvious that the relative humidity measurements can be very useful if the user wants to survey
increased and the temperature decreased. Thus, this is a the environment magnetic field, but the main usage of this
normal behavior because the relative humidity represents magnetometer is to provide the heading indicator to the
the amount of moisture in the air reported to a standard magnetic North Pole, which proves useful if the user travels
value of its saturation value at the same temperature and into the wild and has no other orientation solution.
cold air can hold more moisture than warmed air. The acceleration can give the user some data
Generally the relative humidity depends on both regarding the orientation of the device; hence the device
temperature and atmospheric pressure variations. needs to be fixed into a horizontal position, allowing the
sensors to take good samples. In fig. 10 can be seen some
acceleration measurements done in g (1 g ≈ 9.81 m/ ),
along with its variation when the device was tilled at about
100 ˚ on different axis. The user can know if the device is
fixed onto horizontal position by following the z axis of the
accelerometer. This state is highlighted in the 1st and 4th label
from the fig. 10, and is correlated with the position of the
DAQ from the fig 4.

Fig. 12 Magnetic field density measurements

Fig.12 shows the 3-axis magnetic field

measurements with the device tilled at different angles and
which are marked in 4 labels. With orange is highlighted the
mean value of the magnetic field strength which was
calculated with the following formula shown in equation (3)
Fig. 10 Acceleration measurements [11]:

In fig. 11 there are highlighted the instantaneous |He|=√ (3)

angular velocity measurements given by the slightest
movement of the DAQ. This feature is important because can where Hx, Hy and Hz are the magnetic field densities
provide some useful about the stability of the support where measured on the correspondent axes and the He is the mean
is fixed. In a stationary position the gyroscope does not sense value of the magnetic field strength.
any movement, hence the values are 0 on all 3-axis, but The measurements done with the device show a
when a movement has been sensed, like tilling the device in mean value of about 0.43 G for the Earth’s surface magnetic
this case, the angular velocity is now measured. See both fig. field. This type of measurements are very sensitive and could
10 and 11. be influenced also by other electromagnetic and magnetic
sources (e.g. phones, TVs, light sources, power supply
sources) placed in the near vicinity of the device, but by all
means the measured values are corresponding with other
measurements given in the literature for the Earth’s surface
magnetic field strength [17].
To calculate the compass headings used to
determine the direction of the magnetic North pole using the
same magnetometer, the device must be placed at the Earth’s
surface, and need to be cleared out from any ferrous
materials and also the declination angle specific for each
zone of the Earth must be known. The compass heading can
be determined using only 2 axes, the x and y axes, of the
Earth’s magnetic field and depending of these 2 axes reading
can be calculated with the following set of equations [11]:
Fig. 11 Angular velocity measurements
Direction (y>0) = 90 – [arcTAN ( )] * (4)
The Earth’s magnetic field on the surface was Direction (y<0) = 270 – [arcTAN ( )] * (5)
measured during the indoor-outdoor environmental
measurements, and the results are highlighted in the fig.12. Direction (y=0, x<0) = 180.0 (6)
The values are given in G (gauss), but could be converted Direction (y=0, x>0) = 0.0 (7)
into T (tesla) which is used by the International System of
Units (SI) with the following formula shown in equation (2):

1 G= T (2)
 V. CONCLUSIONS & FUTURE WORK [3] J.K. Lee, A. Christen, Z. Nesic, R. Ketler, “A
This project has proved via experimentation the Mobile Sensor Network to Map CO2 in Urban
reliability of a smart, low-power and wearable system Environments”, 9th International Conference on
designed to monitor, acquire and store environmental Urban Climate jointly with 12th Symposium on the
temperature, humidity, atmospheric pressure, altitude, CO 2 Urban Environment, June 2015.
concentration level. Besides that, it returns the Earth’s [4] http://www.uradmonitor.com, retrieved Nov. 2016.
magnetic field, the North Pole indication, 3-axis [5] Sensirion, “Humidity and Temperature Sensor IC”,
accelerometer measurements and 3-axis angular velocity SHT21 datasheet, Version 3, December 2011.
measurements. [6] CO2 meter, ”COZIR Ultra Low Power Carbon
The device capability has been assessed using a Dioxide Sensor”, COZIR-AJ-5000 datasheet.
testing strategy that included three scenarios: indoor [7] Rae systems, “Theory and operation of the NDIR
measurements, outdoor measurements and indoor- outdoor Sensors”, Technical note TN-169.
measurements. It has shown promising results and a good [8] Bosch, “Digital Pressure sensor”, BMP180
match with the values found in the literature. For example, datasheet, revision 2.5 Apr. 2013.
the CO2 levels indoor were found to be around 1500 ppm. [9] M. Cavcar, “The International Standard
That, according with the literature it is a regular level for this Atmosphere (ISA)”, pp. 4.
kind of environments, but long hours exposure to high levels [10] Honeywell, “3-Axis Digital Compass IC”,
of CO2 might be very dangerous for the human body. That is HMC5883L datasheet, rev E Feb. 2013.
why there is a constant need for precise environmental [11] Honeywell, “Compass heading using
monitoring inside our homes, our working spaces and magnetometers”, Application note AN-203.
moreover our cities. [12] InveSense, “MPU-6000 and MPU-6050 Product
Apart from the basic concepts of smart sensing and Specification”, MPU-6050 datasheet, revision 3.4
data acquisition proven here, a future line of work would Aug. 2013.
mean an Android interface to the hardware. That should [13] ITeadStudion, “Bluetooth to Serial Port Module”,
replace the LabVIEW one and give more portability and HC-05 datasheet, June 2010.
ease-of-access to the end user. [14] Maxim Integrated, “3.3V/5V/Adjustable-Output,
Step-up DC-DC Converters”, MAX756 datasheet,
REFERENCES rev 2.
[15] Atmel, “8-bit AVR Microcontrollers”,
[1] S. Patel, H. Park, P. Bonato, L. Chan, M. Rodgers, ATmega328P datasheet, Nov. 2010.
“A review of wearable sensors and systems with [16] M Schell, D. Inthout, “Demand Control Ventilation
application in rehabilitation”, Journal of Neuro Using CO2”, ASHRAE Journal, Feb. 2001.
Engineering and Rehabilitation, Apr. 2012. [17] International Association of Geomagnetism and
[2] M. Jin, H. Zou, K. Weekly, R. Jia, A.M. Bayen, C. Aeronomy, Working Group V-MOD, “International
J. Spanos, “Environmental Sensing by Wearable Geomagnetic Reference Field: the eleventh
Device for Indoor Activity and Location generation”, Geophysical Journal International, 183,
Estimation”, Industrial Electronics Society, 2014. 1216-1230, (2010).

View publication stats