International Journal of Safety and Security Engineering

Vol. 12, No. 3, June, 2022, pp. 371-379

Journal homepage: http://iieta.org/journals/ijsse

Intelligent Traffic Light Control System with Priority Lane Intervention

Sunardi1*, Anton Yudhana1, Siswaya2, Arsyad Cahya Subrata1
1
Department of Electrical Engineering Universitas Ahmad Dahlan, Yogyakarta 55191, Indonesia
2
Master Program of Informatics Universitas Ahmad Dahlan, Yogyakarta 55191, Indonesia

Corresponding Author Email: [email protected]

https://doi.org/10.18280/ijsse.120311 ABSTRACT

Received: 25 April 2022 The current traffic light system uses a lot of fixed time signals. Fixed time signals cause
Accepted: 29 May 2022 traffic jams in one or several lanes due to traffic congestion. On the other hand, emergency
vehicles cannot get their rights to the maximum due to the density of these vehicles. This
Keywords: study aims to build an intelligent traffic light with a vehicle density sensor device and a
intelligent traffic light, priority intervention, smartphone application to increase the accuracy of light timing and parse traffic density.
IR sensor, Arduino Bluetooth Control (ABC), Furthermore, in an emergency, they can still overcome and accelerate the speed of passing
Arduino Mega 2560 vehicles by manually intervening with the traffic light through a smartphone application
connected wirelessly to the traffic light. The intelligent traffic light is made with Arduino
Bluetooth Control (ABC), Bluetooth HC-05, Arduino Mega 2560, and infrared (IR)
sensor E18-D80NK. The results show that the green light setting is based on the input
from the IR sensor both near sensor (NS) and far sensor (FS) work well. When NS detects
a vehicle with a green light, it increases Y seconds from the default X seconds. If FS
detects a vehicle with a green light, it increases by Z seconds. Settings with the ABC
application can randomly turn on the green light in four lanes according to the will of the
smartphone operator.

1. INTRODUCTION smart cities, including intelligent transportation systems, allow

for a solution to break down congestion, thereby minimizing
In recent decades, urbanization has shown a significant the possibility of accidents and environmental pollution [12].
increasing trend. Around the world, urbanization aims to This is a positive response to the World Health Organization
create and enhance urban areas [1]. The trend of rapid cases, which states that between 20 to 50 million people suffer
urbanization is more specifically shown in developed and from non-fatal injuries and about 1.35 million people die from
developing countries. This case cannot be separated from traffic accidents.
industry development, especially manufactured goods [2], Smart transport systems play an essential role and have
which become the economic magnet. Over time, a higher become one of the main pillars of the smart city concept.
standard of living and higher economic income in urban areas Various studies have been developed related to this smart
has led to an increasing migration trend from rural to urban transport system. Special tourist services are developed in a
areas. Furthermore, greater diversity and quality of work, traveler information system [13-15]. This system includes
supported by superior infrastructure and services in urban information about dynamic signage and in-vehicle
areas, lead to substantial economic and social growth [3]. The notifications to provide real-time information. The
United Nations projects that two-thirds of the world's information and incentives that emerge from the developed
population, or 7 billion people, will live in urban areas [4]. system increase compliance. A system was developed to
This significant development causes an increase in vehicular improve transport information to the public with location
traffic and has excellent potential and has even been noted to tracking and e-ticketing [16-18]. The system developed has
increase congestion, environmental pollution, and accidents succeeded in increasing convenience and providing real-time
[5]. information to commuters directly. In addition, the data is also
Recently, advanced technology such as smart cities has used to study the behavior of commuters to improve services.
become an interesting topic to be developed. The concept that A safety system for drivers was developed using driver
was carried out from 2009-to 2010 offers a more intelligent assistance systems [19]. This system allows the vehicle to
life by adopting an automated service system [6-8]. Due to analyze the surrounding environment using sensors.
massive urbanization, urban life is increasingly challenging Furthermore, the system assists several vehicle operations,
with limited resources and services [9]. The smart city such as adaptive cruise control, lighting, recognition of driver
emerged as a solution in a new dimension of urbanization with fatigue and traffic signs, collision warning, emergency braking,
various digital information service solutions in transportation, in-car navigation systems, and parking assistance. Another
energy, health care, health, and other aspects closely related to technology in the form of a fleet management system can
human life [10]. Smart cities do not grow instantly. The provide automatic route planning and status updates based on
development of digital culture supports the acceptance of location [20, 21]. This system can optimize driver
technology. It is the main trigger for creating the smart city management, planning, and routing to increase efficiency. The
concept [11]. Concerning traffic density due to urbanization,

371
cooperative communication system was developed by the algorithm implemented in the Arduino IDE was developed
utilizing wireless technology [22-24]. This system can to set the time interval for the traffic light based on the status
transmit data to and from vehicles, infrastructure, and of the IR sensor; iv) system prototype is developed to validate
pedestrians. Automatic traffic management systems utilize its effectiveness and functionality.
road-mounted surveillance devices and other sensors to collect The structure of the paper is as follows: Section 2 describes
real-time information. Other controls such as traffic violations related works conducted by previous studies, Section 3
and dangerous incidents can be identified for immediate introduces the traffic light system, which reviews previous
reporting. The data collected is then used for planning and research related to intelligent traffic lights. Section 4 presents
analysis in the future. A system based on the FOG computing the design and fabrication of the developed system. Section 5
paradigm was developed to consider and direct vehicle routes describes the operational procedures for implementing the
[25]. The system can recommend alternative routes that can system by testing its functionality. Finally, Section 6 reflects
reduce congestion levels. the most relevant conclusions and directions for further
The traffic light is a fundamental part of regulating the improvement.
distribution of vehicles on the highway. The traffic light is a
vital instrument to regulate the smooth flow of traffic. So that
proper traffic light management can minimize traffic accidents. 2. RELATED WORKS
The traffic light regulation currently uses a fixed time signal
system [26, 27]. In this traffic light setting, the system cannot To fulfil the facilities of the smart city concept, the smart
determine the duration of the RED (stop) and GREEN (go) traffic light was developed by controlling the traffic light so
signals independently. As a result, congestion often occurs on that it can regulate vehicle traffic automatically and effectively
one or several lanes while the other lanes are empty. adapt to real-time traffic conditions. Wei et al. [28] proposed
Furthermore, officers or traffic police cannot intervene in a reinforcement learning model to control traffic lights. The
traffic lights if they find traffic conditions requiring special method is then tested with a large-scale real-life traffic data
handling. Special handling by the traffic police is usually collection obtained from surveillance cameras. A framework,
carried out when there is a traffic jam, prioritized vehicles such agent, network, and memory palace are needed in this deep
as ambulances, fire engines, and convoys of official vehicles. reinforcement traffic light. All these needs are carried out
Handling is done by officers going down to the middle of the using the Simulation of Urban Mobility (SUMO). However,
intersection to temporarily stop all vehicles from the the problem case used is still the simplified one intersection
intersection, then giving way to the prioritized vehicles. The case. The results obtained will be different in the case of multi-
proposed studies aim to realize a smart city through a smart intersection.
transportation system. These studies focus on traffic Cools et al. [29] proposed a self-organizing traffic light
information systems and route planning. Several studies focus (SOTL) using the more Vehicle Traffic Simulator (moreVTS)
on vehicle control systems. However, the mass control system platform. This method is achieved by logic: if there are one or
involving many vehicles and various special needs, such as the more cars, then the traffic light will set a more extended stop
priority of emergency vehicles to relieve traffic jams, needs signal. Nevertheless, the more cars that join the group, the less
further study. waiting time at traffic lights will be set to be. The test results
Considering the trend of urbanization and the effects of obtained can reduce the average waiting time by half.
congestion that often occurs in traffic, such as suffering from Cao et al. [30] proposed a pheromone-based traffic
air pollution and delayed emergency vehicles, this research is management framework to reduce traffic congestion. The
proposed. This paper presents the design and implementation strategy used is dynamic vehicle routing and traffic light
of intelligent traffic light control that can reduce traffic jams control. In this study, road congestion is predicted by
by adjusting the length of time the traffic lights are on. The representing each vehicle as an agent that stores pheromones
length of time the red and green lights is based on the level of along the way. Next, alternative routes are established by
traffic of passing vehicles. Furthermore, traffic control officers proactive vehicle rerouting based on global distance and local
can adjust the traffic light manually if vehicles need to be pheromone. Meanwhile, the traffic light is controlled through
prioritized to pass. This system facilitates traffic light control an online strategy. The traffic light controllers developed are
using infrared (IR) sensors to detect vehicles and automate the traffic light control not considering downstream traffic (TLC-
change of traffic light lights according to vehicle crowds or NCDT) and traffic light control considering downstream
manual intervention by traffic control officers. Arduino Uno is traffic (TLC-CDT). The experiment was carried out using
integrated with Bluetooth and used as a microcontroller and SUMO and is reported to have shown superior performance in
wireless communication for manual settings. The developed overcoming road congestion.
system is implemented in a prototype built on a traffic light Sajad Mousavi et al. [31] proposed a traffic signal
simulation board with four intersections. The contribution of prediction method based on a deep policy-gradient and value-
this work is the development of an intelligent traffic light function based agent. The experiment was carried out using
control system that can adjust the duration of the stop and run SUMO. The traffic light control agent receives a snapshot of
signals according to the crowd of vehicles. Furthermore, this the current state from the simulator to be used as a control
system is equipped with manual control to prioritize the lanes signal. The agent estimates the highest control value based on
passed by special vehicles, where the operation can be the mapping of direct observations to control signals by a
controlled wirelessly. The proposed system can effectively policy-gradient based agent. The results obtained are reported
break down congestion in one or several paths through the to find a more stable control policy.
simulation. The main contributions of this work are: i) design Li et al. [32] proposed a two-tier method based on a Hybrid
and implementation of traffic light control system; ii) new Genetic Algorithm (HGA) to optimize traffic light
concept for “stop” and “go” signal time interval estimation, management. The upper-level problem minimizes vehicle
which can be intervened manually with wireless support; iii) travel time, while the lower-level problem focuses on

372
achieving network balance. To separate these two levels of based on the type of intersection. The intersections commonly
problems, HGA is integrated with dynamic traffic assignment found in urban areas are intersections with three
(DTA) so that they can be solved sequentially. Optimization approximations (T), four approximations (Cross), and multiple
of traffic light settings also considers dynamically changing approximations. In Indonesia, road driving rules apply traffic
and stable traffic patterns. The results obtained show the rules on the left side of the road. Generally, the traffic light
optimization of efficient traffic light settings. operating cycle in Indonesia is clockwise. Figure 1 shows a
Tajalli et al. [33] assume all vehicles are connected to cross-type traffic light operating cycle scheme with traffic
develop further coordinated speed optimization and traffic rules on the left side of the road that are commonly applied to
light control in urban networks. These vehicles collect data on urban areas in Indonesia.
the infrastructure to find optimal signal timing and vehicle
speed parameters. The experiment was developed based on the
concept of Cell Transmission Model (CTM) network loading
using the Vissim platform. The results obtained are reported to
be able to coordinate signal timing and improve network
performance due to optimized speed.
Miao and Leitner [34] developed a traffic light signal
scheduling as service information applied to Connected and
Automated Vehicles (CAV). Traffic light signal scheduling
considers the longest and fastest waiting time. The traffic light
is controlled adaptively using Adaptive Traffic Light
Controller Using Static Vehicle Deadline (ATC-SD) and
Adaptive Traffic Light Controller Using Dynamic Vehicle
Deadline (ATC-DD). The system is implemented and verified (a)
using SUMO. The results show that the system can guarantee
deadlines and outperform the default method.
Borges et al. [35] proposed traffic light control using
Hierarchical Reinforcement Learning (HRL) and an Options
Framework. HRL is used to establish a sub-policy to maximize
vehicle flow and minimize waiting time. HRL makes it
possible to combine learn abilities and make decisions. The (b)
results show that the proposed method can work better than the
fixed-time traffic light model. Figure 1. Traffic light with traffic rules on the left side of the
However, the experiments that have been described are still road (a) Cross type, and (b) Clockwise cycle
in a simulation framework. Furthermore, the problem used is
adapted to a simple intersection. Therefore, it is necessary to 3.2 Traditional traffic light
carry out a more in-depth study to apply it to hardware then be
applied and tested in a real environment. Traditional traffic light cycle control generally uses a cable
Saad et al. [36] developed a traffic light control system for network that connects each light centrally. The light control
pedestrian crossing areas automatically using PIR sensors. The system on a traditional traffic light is straightforward because
system performs sensing using a PIR motion sensor and an it is only based on a pre-set time. However, this static time
Arduino microcontroller to detect the presence of pedestrians interval cannot consider the volume of passing vehicles, so
and control the crosswalk traffic lights. The developed system time is often wasted. The length of time the lights are not
is equipped with Bluetooth to transmit signals between traffic optimal also causes traffic jams, increasing pollution.
lights. Experiments are implemented in a real environment.
The results obtained are reported to work effectively and 3.3 Smart traffic light
perform tasks according to predetermined criteria. However,
the PIR sensor can only reach a distance of 3m. Furthermore, A traffic light that was developed with an intelligent system
a system was developed specifically to control traffic lights to is a solution to answer various problems that occur in a
assist pedestrians. traditional traffic light. One of the focuses of smart traffic light
research is automatic traffic light control which was developed
to regulate traffic flow by controlling vehicle speed and
3. TRAFFIC LIGHT SYSTEM optimizing traffic light signals to control emergency vehicle
priorities. Research related to automatic traffic light control is
Traffic signals were initiated by John Peake Knight in 1866 summarized in Table 1 to facilitate comparing the models
in London with a police-operated semaphore [37]. This developed based on related works in Section 2.
highway signaling system adopts train signals for the daytime Several studies were conducted to obtain the optimal traffic
and gas-powered red-green lights at night. In 1868, the first light signal time interval. Literatures [28-35] simulated traffic
traffic light was installed on a fixed highway with gas-powered light control with different platforms and methods. Meanwhile,
lights. Until finally, electric-powered traffic lights were [36] develop a prototype of a traffic light signal in pedestrian
installed in London in 1929. crossing areas. This paper builds a prototype to control traffic
lights based on vehicle density at a highway intersection.
3.1 Urban intersection Furthermore, officers can set traffic light signals manually to
provide road signals to emergency vehicles or vehicles that
In general, traffic light operations have a specific cycle have priority.

373
 Table 1. Summary of research on automatic traffic light control

Models Definition
Develop a traffic light control with a deep reinforcement learning model with
Wei et al. [28]
actual traffic data obtained from surveillance cameras. The platform used is
Simulation: SUMO
SUMO 2, which provides traffic light control.
Cools et al. [29] Controlling traffic lights based on current traffic conditions. The platform used is
Simulation: more Vehicle Traffic Simulator (moreVTS) more VTS 2006.
Controlling the duration of a traffic light using a pheromone-based approach. The
Cao et al. [30] control strategies developed are traffic light control not considering downstream
Simulation: SUMO traffic (TLC-NCDT) and traffic light control considering downstream traffic
(TLC-CDT).
Sajad Mousavi et al. [31] Predicting traffic light signals using deep policy-gradient algorithms and value-
Simulation: SUMO function-based methods using visual input data from the simulator.
Li et al. [32] Optimizing traffic light signals related to the coordination of split time, cycle
Simulation: Dynamic Traffic Assignment (DTA) with time, phase sequences, and offsets simultaneously using the HGA algorithm on
Hybrid Genetic Algorithm (HGA) microscopic simulation-based DTA.
Tajalli et al. [33] Develop a traffic light control optimization based on the concept of Cell
Simulation: Vissim Transmission Model (CTM) network loading on the Vissim platform.
Miao and Leitner [34] Develop a traffic light signal scheduling as service information on Connected and
Simulation: SUMO Automated Vehicles (CAV).
Borges et al. [35] Controlling traffic lights using Hierarchical Reinforcement Learning (HRL) and
Simulation: SUMO Options Framework.
Saad et al. [36]
Develop a system to regulate traffic in pedestrian crossing areas using PIR sensors
Prototipe: Arduino microcontroller and Passive Infrared
automatically.
(PIR) sensors
Sunardi et al. (proposed) Develop a light traffic time interval control system based on traffic density,
Prototipe: Arduino microcontroller, PIR sensor, and equipped with a manual intervention system using a mobile device to prioritize
mobile device emergency vehicles.

4. PROPOSED INTELLIGENT TRAFFIC LIGHT Figure 2 shows an intelligent traffic light model system
CONTROL using IR sensors to detect vehicles. Each lane is equipped with
two IR sensors. Each of these IR sensors is defined as a near
Based on the issue of congestion and air pollution resulting sensor (NS) and far sensor (FS) according to the sensor
from the non-optimal time interval of the traffic light signal, distance to the traffic light. The approach of the proposed
this paper proposes the implementation of intelligent traffic intelligent traffic light control system is to detect vehicles at
light control to overcome these problems. Furthermore, the points close to the intersection and points far from the
proposed system is equipped with an automatic controller intersection. The proposed system aims to optimize traffic
connected to a smartphone wirelessly so that officers can light time intervals based on traffic flow density detected using
operate it to provide road access to vehicles that have the right IR sensors. The time interval optimization mechanism in the
to be prioritized. proposed system is based on the pre-set RED signal timing. If
The modeling and implementation of the system are done the primary time RED signal has expired, the vehicle is still
by developing a traffic prototype using a cross intersection detected on the sensor.
with left-hand traffic rules and the traffic light operating cycle
in a clockwise direction. The next stage is to develop software 4.1 Design and fabrication
that is applied to smartphones. The development of the
prototype is done by selecting the sensor material, wireless The system developed consists of three central units: traffic
communication, and controller used, then assembling light control devices, applications on mobile devices, and
electricity and developing algorithms using Arduino IDE. traffic light boards. The traffic light control device consists of
After all the devices are integrated into the prototype board an Arduino Mega as a microcontroller and an IR sensor. This
and the application on the smartphone, the last stage is testing traffic light control device is also equipped with Bluetooth as
to validate the system that has been developed. a communication protocol in the application installed on the
mobile device. Furthermore, the traffic light board was built as
a prototype using boards and LEDs.

4.1.1 Arduino Mega 2560

This system uses a microcontroller as the central core. The
Arduino Mega 2560 board is an Arduino board that uses the
ATmega 2560 microcontroller IC which was chosen as the
microcontroller in this work. This board has relatively many
I/O pins, 54 digital Input/Outputs, 15 of which can be used as
PWM outputs, and 16 analog inputs labeled A0 to A15 as
ADC. Each Analog Pin has a resolution of 10 bits and 4
UARTs. Arduino Mega 2560 is equipped with a 16 MHz
crystal. The function of the Arduino Mega 2560 in this work
Figure 2. Intelligent traffic light model system is to receive an input signal from an infrared object detector

374
sensor (E18-D80NK). There are eight mounted on four sides required a device called a modulator. The modulator is the
at a crossroads. There are two sensors on each side of the road: process of superimposing data on the frequency of the carrier
NS (sensor for a near distance) and FS (sensor for a far signal to the information/message signal so that it can be sent
distance). In addition to receiving an input signal from the to the receiver via certain media (cable or air), usually in the
E18-D80NK, the Arduino Mega 2560 also receives an input form of a sine wave. The modulation of the sine wave will
signal from the HC-05 Bluetooth module, which is sent from convert a baseband message wave into a passband waveform.
the Arduino Bluetooth Control (ABC) application installed on So, this modulator, together with the infrared LED, functions
the android smartphone via Bluetooth to Bluetooth HC-05. to produce modulated light waves, which will be emitted to the
These two input signals will then be processed by the Arduino object continuously.
Mega 2560 to determine the length of time the green light is In the second stage, the receiver converts the modulated
on in the Traffic Light Display. The programming of the light waves into an electrical signal in the form of a
algorithm that is included in the Arduino Mega 2560 is carried photodetector/phototransistor. Suppose an object blocks
out using the Arduino IDE, which can be accessed openly. infrared light. The phototransistor converts the reflected light
from the detected object (light energy) into a current and sends
4.1.2 IR sensor it to the Amplifier. This Amplifier is an IC Op-Amp
The E18-D80NK IR sensor is used in this system. The basic (Integrated Circuit Operational Amplifier). Op-Amp is a
working concept of the object detection sensor is divided into Linear IC that functions as an electrical signal amplifier. Op-
two stages. The first stage is when it functions as a transmitter, Amp consists of transistor components, diodes, resistors, and
and the second stage is when it functions as a receiver. Figure capacitors that are interconnected and integrated to enable it to
3 shows a block diagram of the two stages. produce high gain (gain) over a wide frequency range. The Op-
Amp IC has two inputs (one inverting input and one non-
inverting input) and has one output. The Op-Amp IC also has
two pins for the power supply, namely the positive and
negative pins. The electrical signal that the IC Op-Amp has
amplified is then fed through the output pin to the demodulator.
The demodulator has the opposite function of the modulator
(demodulation), which is getting back data or reading data
from the signal received from the sender. After going through
the clock logic, the received signal will be given to the load as
output.
The back of this IR sensor has a potentiometer (VR) to help
adjust the detection distance. There is a VR tune information
to adjust the distance of detection. So, the basic work of this
sensor is if the object is in front of the sensor and is reached
by the sensor, the sensor circuit output will be logic 0 or low,
which means the sensor detects the object. Conversely,
suppose the object is in a position that is not reached by the
sensor. In that case, the sensor circuit output will be 1 or high,
which means the sensor does not detect the object.
Figure 3. Block diagram of the Infrared Object Detection
4.1.3 Bluetooth module
Sensor Model E18-D80NK-N
The HC-05 Bluetooth module is a TTL level serial
communication converter (UART) into a form of wireless
In the first stage, the transmitter converts the electrical
communication. Bluetooth is used for wireless communication
signal into light waves in the form of a light-emitting diode
protocols for traffic light control. This Bluetooth module can
(LED). The infrared LED continuously emits infrared light to
function as a master or slave. Unlike the other variants, namely
the object to be detected. This section is called the modulated
the HC-06 Bluetooth module which can only function as a
LED. The power or power supply functions to ensure that the
slave. Bluetooth modules HC-05 and HC-06 physically have
voltage inputted to the infrared LED is truly pure and stable
the same shape. The most straightforward distinction is to look
DC. To support these needs, a rectifier regulator is installed.
at the number of pins contained in each module. The HC-05
The rectifier functions to convert the AC signal into a DC
Bluetooth module has a six-pin connector, while the HC-06
signal, while the regulator is a regulation circuit or voltage
has a four-pin connector.
regulator from a power supply so that the fluctuating effect of
Figure 4 describes the wiring connection between the
the input voltage does not affect the output voltage of the
Arduino Mega 2560 and the HC-05 Bluetooth module.
power supply and remains stable. This infrared object
Bluetooth Module HC-05 is a wireless communication module
detection sensor includes a device sensitive to voltage
via Bluetooth that operates at a frequency of 2.4 GHz with the
fluctuations, so it is necessary to have an active component
option of two connectivity modes. Mode 1 (slave mode) acts
that can regulate this output voltage to be stable. So, this
as a data slave/receiver only, while mode 2 (master mode) acts
rectifier regulator is needed to limit the current and voltage to
as a master/transmitter or can also act as a transceiver
remain stable and the infrared LED can produce light at a
(transmitter & receiver). It also has two configuration modes,
constant brightness. The sensor sensitivity level is not
namely AT Mode and Communication Mode. AT Mode is
disturbed. Modulation is the process of laying an information
used for HC-05 module configuration settings. For example, it
signal on a carrier signal. The carrier signal or carrier signal
was changing the Bluetooth password and name. At the same
parameters are varied concerning another (i.e., the modulating
time, the Communication Mode functions for wireless
signal in the form of an information signal). The modulation

375
communication with other devices or devices or can be said to will work based on the instructions given by the mobile device.
be running mode. The interfaces used to access this module If there is no intervention, the system will read the condition
are serial TxD, RxD, Vcc, and GND. There is an LED (built- of the density of vehicles, especially on the path with GREEN
in) as an indicator of Bluetooth connection to other devices light signals. Suppose there is traffic congestion, which is
such as fellow modules, android smartphones, and so on. The detected by the IR sensors. In that case, the system will give
effective range of this module when connected is 10 m, and if additional time on the path with the GREEN light signal,
it exceeds this distance, the quality of connectivity will not be giving longer time on the path of the RED signals. The system
optimal. will provide a default interval time if no traffic density is
detected.

Figure 4. Configuration of Bluetooth serial communication

So, the Bluetooth Module HC-05, which is connected to the

Arduino Mega 2560 microcontroller, functions to receive
command signals from smartphones through the Arduino Figure 5. Concept of the operating cycle with lane
Bluetooth Control (ABC) application that is already installed intervention
in it. Using the features available in ABC, the smartphone will
then send specific command signals to determine/control the
lighting via Bluetooth on the Android Smartphone, which the 5. RESULT AND DISCUSSION
HC-05 Bluetooth module will receive.
The Arduino Mega 2560 board and its supporting
4.1.4 Arduino Bluetooth Control components, such as the HC-05 Bluetooth module and the
Arduino Bluetooth Control (ABC) is an android application E18-D80NK IR sensor, are assembled into a single circuit
that allows creating Bluetooth-based projects or controlling integrated into the traffic light display box. In this system, the
Arduino boards via Bluetooth. Projects are created by adapting ATmega 2560 microcontroller IC is planted with a program so
the new features available in the ABC application. The settings that the system can work according to the original purpose of
section allows project developers to customize the required this device. The result is a prototype intelligent traffic light
applications through a straightforward and intuitive interface. controlled using a smartphone, as shown in Figure 6.
This application will also intelligently remember the
Bluetooth module the user has used and will always try to
connect to the newest module automatically, so there is no
need to select it every time use it. In this work, the ABC
application is used to control the traffic light or determine the
time when the green light is on in such a way, based on input
from smartphone operators, taking into account the current
traffic conditions in real-time.

4.2 System implementation

The concept of the operating cycle and manual intervention

performed by the operator is shown in Figure 5. The system
starts by initializing all the conditions of the sensors and LEDs.
Next, the traffic light is given a default interval time to run the
traffic light cycle. After the system runs with the default
interval time, the system will read the condition whether there
is intervention from the mobile device that is done manually
by the officer or not. If there is an intervention, the traffic light Figure 6. Proposed of intelligent traffic light

376
 Hardware testing includes testing the connectivity of the object detector sensor test is proven by activating the Serial
ABC application installed on the smartphone to the HC-05 Monitor on the Arduino IDE. Serial Monitor will show logic
Bluetooth module and testing eight infrared object detector 0 if the sensor is closed, and vice versa Serial Monitor will
sensors, both NS and FS. Tests were also carried out on the show logic 1 if the sensor is open. The results of the IR sensor
Arduino Mega 2560 board to receive input signals from the test are shown in Table 2.
infrared object detector sensor (E18-D80NK), which
numbered eight, and input signals from the HC-05 Bluetooth Table 2. IR sensor test E18-D80NK object detector
module. The last is to test whether the traffic light props can
display according to the input from the ABC application & the Lane Sensor Treatment Output
object detector IR sensor. 1 NS open close open close 1 0 1 0
FS open open close close 1 1 0 0
5.1 ABC test 2 NS open close open close 1 0 1 0
FS open open close close 1 1 0 0
3 NS open close open close 1 0 1 0
Testing is done by clicking on the ABC application software FS open open close close 1 1 0 0
that has been installed on the smartphone and making sure 4 NS open close open close 1 0 1 0
Bluetooth is active. To connect to Bluetooth, the user must FS open open close close 1 1 0 0
select the appropriate Bluetooth series. The application will
successfully connect to Bluetooth and display a “Connected to After testing the IR sensor by closing the sensor area using
HC-05” notification, as shown in Figure 7a. objects, the output produced is in line with expectations. The
The proof is also done by pressing buttons 1 to 4, as shown resulting output is 100% following the treatment given for
in Figure 7b. Button A to turn on the green light in lane 1, both NS and FS sensors. This test shows that the E18-D80NK
button B to turn on the green light in lane 2, button C to turn IR sensor can be used as the primary sensor for detecting
on the green light in lane 3, and button D to turn on the green vehicle objects in actual implementation.
light in lane 4. The four buttons can function to activate the
green light in lane 4. four lanes of traffic either 5.3 Traffic light test
alternating/adjusting with the push of a button. At the same
time, the other two buttons are used for secondary purposes. This test is carried out in three stages: i) testing by default
Green signal control manually through the ABC application is when there is no input from the ABC application and the IR
carried out with the following pseudo-code. sensor; ii) testing when there is input from ABC; iii) testing
when there is input from the IR sensor.
1 void serialEvent2() Testing when by default will form a loop starting from cycle
2 {
1 to cycle eight, and after that, it returns to cycle 1. And so on
3 while (Serial2.available()) {
4 char inChar = (char)Serial2.read(); with each predetermined time during programming. The
5 Serial.print(inChar); measure of the success of this first stage of testing is the cycle
6 if (inChar == '1'){counter =0; Serial.println ("Lane 1");} of changing the green light in a clockwise direction. There is a
7 if (inChar == '2'){counter=20; Serial.println ("Lane 2");} sequence of alternating lights starting from lane 1 to lane 4.
8 if (inChar == '3'){counter=40; Serial.println ("Lane 3");} The second stage of testing is when there is an input / given
9 if (inChar == '4'){counter=60; Serial.println ("Lane 4");} input from ABC through the buttons, namely buttons one, two,
10 } three, or four, as follows: i) when the A button is pressed, lane
11 } one green lights up, and lanes two, three, and four all turn red;
ii) when the D button is pressed, lane four green lights up, and
lanes one, two, and three all turn red; iii) when the B button is
pressed, lane two green lights up, and lanes one, three, and four
all turn red; iv) when the C button is pressed, lane three green
lights up, and lanes one, two, and four all turn red.
The measure of the test's success in this second stage is
when the user as a controller, for example, a traffic police
officer, presses a button either randomly or sequentially. The
GREEN light on the props lights up according to the button
pressed. The third testing stage is when there is input from the
IR sensor. As previously explained, two sensors are used in
each lane, namely NS and FS. So, for the four lanes, there are
eight sensors in all. According to the program design that is
(a) (b) made for NS, if it detects a vehicle, the green light, which
defaults to lit for X seconds, will be added for Y seconds so
Figure 7. The display of the ABC application (a) connected that the light turns on for XY seconds. Likewise, for FS, if it
to the traffic light system via Bluetooth HC-05 series, and (b) detects a vehicle, the green light that defaults to lit for X
the button to manually activate the GREEN signal traffic seconds will be added for Z seconds to turn the light to XZ
light seconds. The success parameter of this third stage is the
addition of the green light time according to the sensor that
5.2 IR E18-D80NK sensor test detects the vehicle. Based on the three stages of testing carried
out, it can be concluded that the traffic light props can display
The test in this section is carried out by alternating blocking according to the input given, either by default, coming from
sensors, both NS and FS. The parameter for the success of the the ABC application or input from the IR sensor.

377
6. CONCLUSION sustainable smart living. Habitat International, 108:
102312.
Intelligent traffic light control has been developed to https://doi.org/10.1016/j.habitatint.2021.102312
provide manual intervention based on the lane when an [8] Araral, E. (2020). Why do cities adopt smart
emergency vehicle is passing. The system is built with the technologies? Contingency theory and evidence from the
Arduino Mega 2560 microcontroller programmed through the United States. Cities, 106: 102873.
Arduino IDE. Through this software, Arduino is programmed https://doi.org/10.1016/j.cities.2020.102873
to perform functions embedded through programming syntax, [9] Kirimtat, A., Krejcar, O., Kertesz, A., Tasgetiren, M. F.
and an ABC application is installed on a smartphone to control (2020). Future trends and current state of smart city
the intelligent traffic light system via Bluetooth. The results concepts: A survey. IEEE Access, 8: 86448-86467.
showed that the green light setting, which rotates clockwise https://doi.org/10.1109/ACCESS.2020.2992441
with each green light on each lane for X seconds, can work [10] Sivrikaya, F., Ben-Sassi, N., Dang, X.T., Görür, O.C.,
well. The input from the infrared object detection sensor, both Kuster, C. (2019). Internet of smart city objects: A
NS and FS, can be responded to well by the program distributed framework for service discovery and
created/uploaded to the Atmega 2560 microcontroller chip. composition. IEEE Access, 7: 14434-14454.
When the NS detects a vehicle, the green light will increase Y https://doi.org/10.1109/ACCESS.2019.2893340
seconds from the default X seconds. Likewise, if the FS [11] Kashef, M., Visvizi, A., Troisi, O. (2021). Smart city as
detects a vehicle, the green light will increase by Z seconds a smart service system: Human-computer interaction and
from the default X seconds. Likewise, the ABC application smart city surveillance systems. Computers in Human
software settings can randomly turn on the green light in the Behavior, 124: 106923.
four existing lanes, according to the wishes of the smartphone https://doi.org/10.1016/j.chb.2021.106923
operator. So, both the eight infrared object detection sensors [12] de Oliveira, L.F.P., Manera, L.T., Da Luz, P.D.G. (2020).
and the ABC application software can function to regulate Development of a smart traffic light control system with
traffic flow at crossroads or the existing four-lane roads adjust real-time monitoring. IEEE Internet of Things Journal,
to the density of passing vehicles. The limitation of this 8(5): 3384-3393.
research is that it is still a prototype. The future research that https://doi.org/10.1109/JIOT.2020.3022392
will be carried out is to implement it in real areas. Testing the [13] Moreno, A., Salaberria, I., López-de-Ipiña, D. (2015).
IR sensor for temperature changes also needs to be tested in- New approaches in user services development for
depth. multimodal trip planning. Intelligent Transport Systems:
Technologies and Applications, (313).
https://doi.org/10.1002/9781118894774.ch16
ACKNOWLEDGMENT [14] Bauer, J., Bedsole, L., Snyder, K., Neuner, M., Smith,
M.C. (2018). Expanding traveler choices through the use
This work was supported by Universitas Ahmad Dahlan of incentives: A compendium of examples (No. FHWA-
(Grant No.: R/124/B.6/III/2022). HOP-18-071). United States. Federal Highway
Administration.
[15] Paz, A., Peeta, S. (2009). Paradigms to deploy a
REFERENCES behavior-consistent approach for information-based real-
time traffic routing. Networks and Spatial Economics,
[1] Szabo, C.P. (2018). Urbanization and mental health: A 9(2): 217-241. https://doi.org/10.1007/s11067-008-
developing world perspective. Current Opinion in 9077-4
Psychiatry, 31(3): 256-257. [16] Mauri, A., Khemmar, R., Decoux, B., et al. (2020). Deep
https://doi.org/10.1097/YCO.0000000000000414 learning for real-time 3D multi-object detection,
[2] Gollin, D., Jedwab, R., Vollrath, D. (2016). Urbanization localisation, and tracking: Application to smart mobility.
with and without industrialization. Journal of Economic Sensors, 20(2): 532. https://doi.org/10.3390/s20020532
Growth, 21(1): 35-70. https://doi.org/10.1007/s10887- [17] Evans, G., Guo, A.W., Blythe, P., Burden, M. (2015).
015-9121-4 Integrated smartcard solutions: Do people want one card
[3] Rodriguez-Hernandez, M.A., Gomez-Sacristan, A., for all their services? Transportation Planning and
Gomez-Cuadrado, D. (2019). SimulCity: Planning Technology, 38(5): 534-551.
communications in smart cities. IEEE Access, 7: 46870- https://doi.org/10.1080/03081060.2015.1039233
46884. https://doi.org/10.1109/ACCESS.2019.2909322 [18] Caulfield, B., O'Mahony, M. (2007). An examination of
[4] Ritchie, H., Roser, M. (2018). Urbanization. Our World the public transport information requirements of users.
in Data. IEEE Transactions on Intelligent Transportation Systems,
[5] Qi, L., Zhou, M., Luan, W. (2015). Emergency traffic- 8(1): 21-30. https://doi.org/10.1109/TITS.2006.888620
light control system design for intersections subject to [19] Pérez, J., Gonzalez, D., Milanés, V. (2015). Vehicle
accidents. IEEE Transactions on Intelligent control in ADAS applications. Intelligent Transport
Transportation Systems, 17(1): 170-183. Systems: Technologies and Applications, Chapter 11,
https://doi.org/10.1109/TITS.2015.2466073 206-219. https://doi.org/10.1002/9781118894774.ch11
[6] Lv, Z., Chen, D., Li, J. (2021). Novel system design and [20] McLeod, S., Scheurer, J., Curtis, C. (2017). Urban public
implementation for the smart city vertical market. IEEE transport: Planning principles and emerging practice.
Communications Magazine, 59(4): 126-131. Journal of Planning Literature, 32(3): 223-239.
https://doi.org/10.1109/MCOM.001.2000945 https://doi.org/10.1177%2F0885412217693570
[7] Han, M.J.N., Kim, M.J. (2021). A critical review of the [21] Hu, Y.C., Chiu, Y.J., Hsu, C.S., Chang, Y.Y. (2015).
smart city in relation to citizen adoption towards Identifying key factors for introducing GPS-based fleet

378
 management systems to the logistics industry. [30] Cao, Z., Jiang, S., Zhang, J., Guo, H. (2016). A unified
Mathematical Problems in Engineering, 2015: 413203. framework for vehicle rerouting and traffic light control
https://doi.org/10.1155/2015/413203 to reduce traffic congestion. IEEE Transactions on
[22] Knorr, F., Baselt, D., Schreckenberg, M., Mauve, M. Intelligent Transportation Systems, 18(7): 1958-1973.
(2012). Reducing traffic jams via VANETs. IEEE https://doi.org/10.1109/TITS.2016.2613997
Transactions on Vehicular Technology, 61(8): 3490- [31] Sajad Mousavi, S., Schukat, M., Howley, E. (2017).
3498. https://doi.org/10.1109/TVT.2012.2209690 Traffic Light Control Using Deep Policy-Gradient and
[23] Mandžuka, S., Ivanjko, E., Vujić, M., Škorput, P., Value-Function Based Reinforcement Learning. arXiv e-
Gregurić, M. (2015). The use of cooperative its in urban prints, arXiv-1704.
traffic management. Intelligent Transport Systems: [32] Li, Z., Shahidehpour, M., Bahramirad, S., Khodaei, A.
Technologies and Applications, Chapter 14, 272-288. (2016). Optimizing traffic signal settings in smart cities.
https://doi.org/10.1002/9781118894774.ch14 IEEE Transactions on Smart Grid, 8(5): 2382-2393.
[24] Eloranta, P., Sukuvaara, T. (2015). Wireless https://doi.org/10.1109/TSG.2016.2526032
communications in vehicular environments. Intelligent [33] Tajalli, M., Mehrabipour, M., Hajbabaie, A. (2020).
Transport Systems: Technologies and Applications, Network-level coordinated speed optimization and
Chapter 3, 37-56. traffic light control for connected and automated vehicles.
https://doi.org/10.1002/9781118894774.ch3 IEEE Transactions on Intelligent Transportation Systems,
[25] Brennand, C.A., Boukerche, A., Meneguette, R., Villas, 22(11): 6748-6759.
L.A. (2017). A novel urban traffic management https://doi.org/10.1109/TITS.2020.2994468
mechanism based on fog. In 2017 IEEE symposium on [34] Miao, L., Leitner, D. (2021). Adaptive traffic light
computers and communications (ISCC), Heraklion, control with quality-of-service provisioning for
Greece, pp. 377-382. connected and automated vehicles at isolated
https://doi.org/10.1109/ISCC.2017.8024559 intersections. IEEE Access, 9: 39897-39909.
[26] Miller, A.J. (1963). Settings for fixed-cycle traffic https://doi.org/10.1109/ACCESS.2021.3064310
signals. Journal of the Operational Research Society, [35] Borges, D.F., Leite, J.P.R., Moreira, E.M., Carpinteiro,
14(4): 373-386. https://doi.org/10.1057/jors.1963.61 O.A. (2021). Traffic light control using hierarchical
[27] Webster, F.V. (1958). Traffic signal settings (No. 39). reinforcement learning and options framework. IEEE
[28] Wei, H., Zheng, G., Yao, H., Li, Z. (2018). Intellilight: A Access, 9: 99155-99165.
reinforcement learning approach for intelligent traffic https://doi.org/10.1109/ACCESS.2021.3096666
light control. In Proceedings of the 24th ACM SIGKDD [36] Saad, W.K., Hashim, Y., Jabbar, W.A. (2020). Design
International Conference on Knowledge Discovery & and implementation of portable smart wireless pedestrian
Data Mining, London United, Kingdom, pp. 2496-2505. crossing control system. IEEE Access, 8: 106109-
https://doi.org/10.1145/3219819.3220096 106120.
[29] Cools, S.B., Gershenson, C., D’Hooghe, B. (2013). Self- https://doi.org/10.1109/ACCESS.2020.3000014
organizing traffic lights: A realistic simulation. In [37] BBC. The man who gave us traffic lights. (2009).
Advances in Applied Self-Organizing Systems, pp. 45- https://www.bbc.co.uk/nottingham/content/articles/2009
55. https://doi.org/10.1007/978-1-4471-5113-5_3 /07/16/john_peake_knight_traffic_lights_feature.shtml#
:~:text=.

379