Optimisation of Distance Measurement in Autonomous Vehicle using Ultrasonic and LIDAR Sensors

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 05 | May 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2072
Optimisation of Distance Measurement in Autonomous Vehicle using
Ultrasonic and LIDAR Sensors
Sandeep Chowdhry1
1Engineering Consultantant and Trainer, Chandigarh, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – Autonomous Vehicles are sharply gaining
popularity in the engineering and technology sector. This
study aims to optimise the distance measurement of an
Autonomous Vehicle using an ultrasonic sensor and LIDAR
sensor for education purposes. An experiment is performed to
find an indoor operating range of an ultrasonic and LIDAR
sensor. In addition, Full-Factorial design is used to
experimentally find the effect of angle, distance and object
shape on the sensor measurement error rate and the object
detection rate. The result shows that an ultrasonic and LIDAR
sensor has indoor operating ranges different from the
manufacturer’s specifications. Both sensors are sensitive to
angle, distance and object shape. It is concluded that the
ultrasonic and LIDAR sensor has indoor operating ranges of
2000 mm and 250 mm. The ultrasonic sensor has minimum
measurement error for cylinder shape objects, 0° angle and a
distance value of 500 mm. It has a maximum object detection
rate for cuboid shape objects, an angle value of 0° and a
distance value of 500 mm, respectively. LIDAR sensor has a
minimum measurement error for cuboid shape objects, 0°
angle and a distance value of 250 mm. It has a maximum
object detection rate for cuboid shape objects,0° angle value
and distance value of 250 mm, respectively.
Key Words: Autonomous Vehicles, Full Factorial Design,
Ultrasonic sensor, LIDAR sensor, Distance measurement.
1. INTRODUCTION
The autonomous vehicles (AV) market is expected to rise by
about 59% between 2020 and 2023 [1]. AV uses different
sensors to interact with the environment to automate the
process. The software's used to increase the productivity of
the automated technology. AVs are used in private vehicles,
transportation and the farming industry. LIDAR sensors are
used in AV to detect objects 200m to 300m away and emit
short pulses to measure more than a million points per
second. It creates three-dimensional maps of the
environment to visualise theobjectaroundthe environment.
They assist in covering a 360° view and helpcreatea 3Dmap
to have a clear sight of what is happening around the vehicle
and make it capable of "seeing "[2]. Google AV company
Waymo started designing and producing their LIDAR
sensors to reduce costs and develop AV for the massmarket.
Tesla AV can help auto steer, auto park,autolanechangeand
fully self-drive long distances on highways for a significant
period without human assistance.
However, Tesla is not using LIDAR sensors costing 75000
dollars per unit. Instead, it uses ultrasonic sensors for car
detection and collisionprevention[3].Ultrasonicsensors are
a cheap alternative to LIDAR sensors for object detection
systems and algorithms in AV. It sends ultrasonic impulses
that are then reflected by the obstacle. Therefore,ithelpsAV
to perform according to the barrier. Ultrasonic sensors can
range up to 5.5m and have limitations such as difficulty in
detecting objects going at a fast speed. They are vulnerable
to jamming and spoofing attacks leaving the sensor
physically unable to function and creating false positives. It
could lead to a potential incident without user
supervision[4]. The measurement of the ultrasonicsensor is
sensitive to temperature and the angle of the target. In
addition, some materials are more absorbent than others,
and these will reflect less ultrasound [5]. Therefore, it
complicates measuring the distance with an ultrasonic
sensor alone.
On the other hand, the LIDAR sensor cannot recognise
transparent objects. So it is advantageous to use an
ultrasonic sensor and LIDAR sensor to detect transparent
objects [6]. Instead of taking 100 measurements per
measuring location in an ultrasonic sensor, 20
measurements per measuring location create a relatively
good environment occupancy grid utilised for robot
navigation tasks [7]. The operating distance commonly
stated by manufacturers of air ultrasound range finding
modules and devices can be very misleading. It should be
estimated experimentally [8].
The literature review indicates that using ultrasonic and
LIDAR sensors for distance measurement is more
advantageous than utilising either sensor alone. Therefore,
this study aims to optimise the distance measurement of an
AV using an ultrasonic sensor (HC-SR04) and LIDAR sensor
(VL53L0X) for education purposes. The main objectives of
this research are 1) To experimentally find an effective
indoor operating range of an ultrasonic sensor and LIDAR
sensor; 2) To experimentally find the effect of an angle,
distance and object shapes on the error rate and detection
rate of an ultrasonic sensor and LIDAR sensor. The study
intends to contribute to the literature on optimising the
distance measurement of an AV using ultrasonic and LIDAR
sensors.

2. EQUIPMENT
2.1 Autonomous Vehicle
The AV body (Fig. 1) is built with upper and lower plastic
sheets. A geared motor is attached to each of the fourwheels
mounted on the side of the plastic sheets. An ultrasonic
sensor is mounted on the breadboard. LIDAR sensor is
mounted on the front side of the AV between the lower and
upper plastic sheet. The LIDAR sensor is located on the
centre line between the transmitter and receiver of the
ultrasonic sensor. The breadboard is powered with an
Arduino Uno microcontroller through the laptop. Ultrasonic
sensor Trig and Echo pins are connected to pins 10 and 9 on
Arduino. LIDAR sensor’s SDA and SCL pins are connected to
analogue pins A4 and A5 on Arduino.
Fig -1: Experiment setup
2.2 Arduino Uno Microcontroller
Fig -2: Arduino Uno microcontroller board
Arduino Uno microcontroller(Fig.2) technical specifications
are Atmega328P; operating voltage:5V; Input voltage
(recommended): 7-12V; Input V. (limit): >6, <20V; PWM
Digital I/O Pins:6; Analogue Input pins: 6; DC current per
I/O: 20mA; DC current for 3.3V: 50Ma; Flash Memory:
32KB,0.5KB used by loader; Clock speed: 16MHz; Length,
width, weight: 68.6mm,53.4mm,25g respectively.
2.3 Ultrasonic Sensor
Fig -3: Ultrasonic sensor (HC-SR04)
Ultrasonic sensor(Fig.3) technical specificationsareCurrent
voltage (V): DC 5V; Ground volatge (G): 0V; Workingcurrent
(C): 15mA; Working frequency (F): 40KHz; Range
(Max/Min): 400 cm/2 cm; Angle measure: 5-15 degree;
Trigger signal: 10uS TTL pulse; Echo signal: Depend on max
range of TTL; Dimensions: 45 mm x 20 mm x 15 mm
respectively.
2.4 LIDAR Sensor
Fig -4: LIDAR Sensor (VL53L0X)
LIDAR sensor’s (Fig. 4) technical specifications are Package:
Optical LGA12; Size: 4.40 x 2.40 x 1.00 mm; Operating
voltage: 2.6 to 3.5V; Operating tempertaure: -20 to 70°C;
Infrared emitter: 940 nm; I2C: Up to 400 kHz (FAST mode)
serial bus Address: 0X52 respectively.
2.5 Three dimensional objects
Three 3-dimensional objects used are Cuboid: 156 x 3 x 213
mm; Cylinder: Diameter 85 mm and height 167 mm; Cone:
base diameter 113 mm and vertical height 242 mm
respectively.

2.6 Arduino Program
Fig -5: Flowchart of Arduino program
3. EXPERIMENT
3.1 Parameters, Levels and Responses
Table 1 shows the three-level settings of the parameters
such as shape, angle and distance. The sensors'
measurement error rate and object detection rate are
selected as the response.
Table -1: Process parameter levels
Parameter Level 1 Level 2 Level 3
Shape Cuboid Cylinder Cone
Angle (deg) 0° 5° 10°
Distance (mm) 250 500 1000
3.2 Design of Experiment
Initially, the Minitab 2019 software is used to design the
random run order for the full-factorial experiment. The
parameters are chosen as categorical in nature. It consistsof
27 points (Run 1-27), one replicate and one block, as shown
in Table 2.
Table -2: Full Factorial Design of three factors with three
levels
Run
Order Shape Angle (deg)
Distance
(mm)
1 Cylinder 0° 500
2 Cone 5° 500
3 Cone 10° 1000
4 Cylinder 5° 250
5 Cuboid 5° 250
6 Cylinder 0° 1000
7 Cylinder 10° 500
8 Cone 10° 250
9 Cuboid 10° 500
10 Cuboid 10° 250
11 Cuboid 10° 1000
12 Cone 0° 1000
13 Cuboid 0° 500
14 Cylinder 0° 250
15 Cylinder 10° 1000
16 Cuboid 5° 1000
17 Cone 5° 250
19 Cone 5° 1000
20 Cone 10° 500
21 Cylinder 5° 500
22 Cuboid 5° 500
24 Cuboid 0° 1000
25 Cuboid 0° 250
26 Cone 0° 500
27 Cone 0° 250
2.6 Procedure
First, the cuboid shape object is placed at an angle of 0° at a
distance of 250 mm, 500 mm, 1000 mm, 2000 mm and 3000
mm, respectively. Twenty measurementsofboththesensors
are recorded at each distance value.

Second, the object is placed as per the shape, angle, and
distance's first run order values, as shown in Table 2. Ten
distance measurements are recorded for boththeUltrasonic
sensor and LIDAR sensor, as shown in Fig. 6. During the
experiment, it is noted that when the LIDAR sensor cannot
detect an object, it offers an enormous value of 8190 mm or
8191mm. Similarly, when an Ultrasonicsensorcannotdetect
an object, it shows an immense value of 11652 mm Etc. In
such cases, it is considered that the sensor with a maximum
measurement error rate and no detection. Among the ten
recorded values for each measurement location, the mode
value is used as the sensor's reading. The maximum error
reading is recorded as the measurement in the absence of
mode. It is used to calculate the sensor error rate.
Similarly, out of ten recorded measurements, if the
enormous value is recorded six times, the sensor detection
rate is calculated as 40%. Again, the response values are
recorded for the ramining run orders as shown in Table 2.
Afterwards, Minitab software is used to analyse the full-
factorial design.
Fig -6: Serial monitor sensor measurement dispaly
4. RESULTS
Fig -7: Sensors distance measurement at 250 mm
The graph in Fig. 7 shows that for twenty measurements of
the exact measurement location, an ultrasonic sensor can
measure the distance of 250 mm with a measurement error
of about 10%. On the other hand, the LIDAR sensor has a
measurement error of about 80%.
Fig -8: Sensors distance measurement at 500mm
measure the distance of 500 mm with a measurement error
below 10%. On the other hand, the LIDAR sensor has a

measurethe distance of 1000mm with a measurementerror
below 5%. On the other hand, the LIDAR sensor has a
The graph in Fig. 10 shows that an ultrasonic sensor can
measure the distance of 2000 mm with a 0% measurement
error for twenty measurements of the exact location. On the
other hand, the LIDAR sensor has a measurement error of
about 100%.
-20%
0%
20%
40%
60%
80%
100%
1 2 3 4 5 6 7 8 9 1011121314151617181920
Measured distance 3000 mm
% Error US % Error LIDAR
accurately measure the distance of 3000 mm for the first
seven readings. It has a measurement error of about 60% for
the remaining readings. On the other hand, the LIDARsensor
has a measurement error above 95%.
Fig -12: Sensors detection rate and error rate
The graph in Fig. 12 shows that the ultrasonic sensor has an
almost negligible measurement error for a distance up to
2000 mm and hasa 100% measurementerroratadistanceof
3000 mm. Second, the LIDAR sensor hasan almost negligible
measurement error up to the distance of 250 mm and has a
100% measurement error for the other distance values.
Third, the ultrasonicsensor hasa 100% object detectionrate
for distances up to 2000 mm and around a 40% object
detection rate for distance values of 3000 mm. Fourth, the
LIDAR sensor has objectdetectionratesof100%atadistance
of 250 mm, approximately 75% at a distance of 500 mm,
about 25% at a distance of 1000 mm and below 5% for
remaining distance values.
Fig -13: Parameters and Ultrasonic sensor error rate
The graph in Fig. 13 shows that an ultrasonic sensor has
minimum measurement error for cylinder shape objects, 0°
angle and a distance value of 500 mm. It has maximum
measurement error for cone shape objects,10° angle value

Fig -14: Parameters and LIDAR sensor error rate
The graph in Fig. 14 shows that the LIDAR sensor has a
minimum measurement error for cuboid shape objects, 0°
Fig -15: Parameters and Ultrasonic sensor detection rate
The graph in Fig. 15 shows that an ultrasonic sensor has a
minimum object detection rate for cone shape objects, 10°
Fig -16: Parameters and LIDAR sensor detection rate
The graph in Fig. 16 shows that the LIDAR sensor has a
minimum object detection rate for cone-shaped objects, 10°
object detection rate for cuboid shape objects, 0° angle value
Fig -17: Optimal settings of Ultrasonic sensor
The graph in Fig. 17 shows that in an ultrasonic sensor, to
maximise the object detection rate and minimise the
measurement error rate, the optimal settings are anobjectof
cuboidshape, angle value of 0° anddistancevalueof500mm,
respectively.
Fig -18: Optimal settings of LIDAR sensor

The graph in Fig. 18 shows that in the LIDAR sensor, to
maximise the object detection rate and minimise the
measurement error rate, the optimal settings are anobjectof
cuboidshape, angle value of 5° anddistancevalueof500mm,
respectively.
5. DISCUSSION
First, for an ultrasonic sensor, an effective indoor operating
range with minimum measurement error and maximum
detection rateis 2000 mm (Fig. 12). However,aftertheinitial
seven accurate measurement readings at a distance value of
3000 mm, the ultrasonic sensor starts giving a measurement
error of above 60% (Fig. 11). It might be possible for an
ultrasonic sensor to provide an accurate distance
measurement reading for a more significant size object at a
distance of 3000 mm. Alternatively, the use of two ultrasonic
sensors in front of the AV might reduce the measurement
error and maximise the detection rate of the sensor at a
distance value of 3000 mm or more. This study agrees with
[8] to experimentally estimate the operating distance of air
ultrasound range finding modules and devices. Second, for a
LIDAR sensor, the indoor operating range with minimum
measurement error and maximum object detection rate is
250 mm (Fig. 12). However, the LIDAR sensor has around a
75% detection rate at a distance value of 500 mm and
approximatelya 25% objectdetectionrateatadistancevalue
of 1000 mm. It might be possible to use a LIDAR sensor
effectivelyabove 250 mm to increase anobjectdetectionrate
by making it rotate within an angle range continuously in
front of an AV with the help of a motor rather than keeping it
stationary. It might be possible for the LIDAR sensor to
measure more significant size objects with minor
measurement error at distance values above 250 mm. Third,
an ultrasonic sensor measurement error rate and object
detection rate are sensitive to the object's shape, angle and
distance, as shown in Fig. 13 and Fig. 15. In addition, the
graph in Fig. 17 indicates that the optimal settings for an
ultrasonic sensor are objects of cuboid shape, angle value of
0° and distance value of 500 mm, respectively. The finding of
the optimal settings of an ultrasonic sensor for the cuboid
body is in agreement with [9]. It indicates that morethanone
ultrasonic sensor may be used on an AV to increase distance
measurement reliability with an ultrasonic sensor for
different shapes, angles, and distances. Fourth, the LIDAR
sensor's measurement error rate and object detection rate
are sensitive to the object's shape, angle and distance, as
shown in Fig. 14 and Fig. 16.
In addition, the graph in Fig. 18 shows that the optimal
settings for the LIDAR sensor are an object of cuboid shape,
angle value of 5° and distance value of 500 mm, respectively.
It shows that the LIDAR sensor and an ultrasonic sensor
increase the object detection zone in front of the AV. It
indicates that both sensors should be used in conjunction to
improve the reliability and accuracy of the distance
measurement and an object detection with them rather than
using them alone. This finding agrees with [10] that a more
reliable car operation is achieved by using data from sensors
after sensor fusion.
6. CONCLUSIONS
This study aims to optimise the distance measurement of an
AV using an ultrasonic sensor (HC-SR04) and LIDAR sensor
(VL53L0X). The main objectives of this research are 1) To
experimentally find an effective indooroperating rangeofan
ultrasonic sensor and LIDAR sensor; 2) To experimentally
find the effect of an angle, distance and object shapes on the
error rate and detection rate of an ultrasonic sensor and
LIDAR sensor. First, the findings are foran ultrasonic sensor,
an effective indoor operating range with minimum
measurement error, and a maximum detection rate is 2000
mm. Second, for a LIDAR sensor, the indoor operating range
with minimum measurement error and maximum object
detection rate is 250 mm. Third, an ultrasonic sensor has
minimum measurement error for cylinder shape objects, 0°
and distance value of 250 mm, respectively. Fourth, an
ultrasonic sensor has a minimum object detection rate for
cone-shaped objects, 10° angle and a distance value of 250
mm. It has a maximum object detection rateforcuboidshape
objects, 0° angle value and distance value of 500 mm,
respectively. Fifth, the LIDAR sensor has a minimum
measurement error for cuboid shape objects, 0° angle and a
distance value of 250 mm. It has maximum measurement
error for cone shape objects,10° angle value and distance
value of 1000 mm, respectively. Sixth, theLIDARsensorhasa
minimum object detection rate for cone-shaped objects, 10°
and distance value of 250 mm, respectively. The limitation of
this research is thatall measurements arerecordedwhilethe
AV is stationery. The suggestion for further research is to
perform the distance measurement with moving AV and use
the data from the sensors after sensor fusion to detect the
objects and measure distance.
REFERENCES
[1] https://www.globenewswire.com/news-
release/2020/05/20/2036203/0/en/Global-
Autonomous-Cars-Market-2020-to-2030-COVID-19-
Growth-and-Change.html M. Young, The Technical
Writer’s Handbook. Mill Valley, CA: University Science,
1989.
[2] J. Hecht, 2018, “Lidar for Self-driving Cars,” Opt.
Photonics News, 29(1), pp. 26–33.K. Elissa, “Title of
paper if known,” unpublished.
[3] T. Khan Mohd, R. Ayala, Sensors in Autonomous
Vehicles: A Survey, Journal of Autonomous Vehiclesand
Systems, July 2021,Vol. 1, pp 1-12.

[4] C. Yan, W. Xu, and J. Liu, 2016, “Can You Trust
Autonomous Vehicles: Contactless Attacks Against
Sensors of Self-driving Vehicle,” Def Con, 24(8), p. 109.
[5] Michal Kelemen , Ivan Virgala , Tatiana Kelemenová ,
Ľubica Miková , Peter Frankovský , Tomáš Lipták , Milan
Lörinc , Distance Measurement via Using of Ultrasonic
Sensor, Journal of Automation and Control, 2015, Vol. 3,
No. 3, 71-74.
[6] T. Zhang∗ , Z. J. Chong∗ , B. Qin∗ , J. G. M. Fu† , S.
Pendleton∗ , M. H. Ang Jr.,SensorFusionforLocalization,
Mapping and Navigation in an Indoor Environment, 7th
IEEE International Conference Humanoid,
Nanotechnology, Information Technology
Communication and Control, Environment and
Management (HNICEM) The Institute of Electrical and
Electronics Engineers Inc. (IEEE) Philippine Section 12-
16 November 2014 Hotel Centro, Puerto Princesa,
Palawan, Philippines.
[7] J. Hanzel , M. KĐúþik , L. Jurišica , A. Vitko , Range finder
models for mobilerobots,ProcediaEngineering48(2012
) pp 189 – 198.
[8] O. S. Sonbul , A. N. Kalashnikov, Determining The
Operating Distance of Air Ultrasound Range Finders:
Calculations and Experiments, International Journal of
Computing, 13(2) 2014, 125-131.
[9] Anuruk, The Study and ComparisontheResultsofObject
Height Measurement with Different Shape using
Ultrasonic Sensor, Engineering Transactions, Vol. 23,
No.1 (46) Jan-Jun 2020.
[10] Muhammad Irfan Haider Jilani, Topic 02: Sensor Fusion
Techniques for Autonomous Driving Applications
BIOGRAPHIES
Er. Sandeep Chowdhry has done
Bachelor of Engineering in
Mechanical Engineering with a
Specialisation in Manufacturing
Engineering from S.L.I.E.T. Punjab,
India and is interested in solving
industrial problems and imparting
training to the professionals in the
industry.(studentsuniversityonline
@gmail.com)

Optimisation of Distance Measurement in Autonomous Vehicle using Ultrasonic and LIDAR Sensors

More Related Content

Similar to Optimisation of Distance Measurement in Autonomous Vehicle using Ultrasonic and LIDAR Sensors (20)

More from IRJET Journal (20)

Recently uploaded (20)

Optimisation of Distance Measurement in Autonomous Vehicle using Ultrasonic and LIDAR Sensors