2019 Sixth International Conference on Internet of Things: Systems, Management and Security (IOTSMS)

Smart Traffic Light Controller System

L. F. P. Oliveira, L. T. Manera, P. D. G. Luz
Dept. of Semiconductor, Instrumentation and Photonic (DSIF)
Univesity of Campinas (UNICAMP)
Campinas, Brazil
Email: [email protected], [email protected],
[email protected]

Abstract—The constant increase in the number of vehicles, the In addition to IoT devices uses, another key agent for
lack of road traffic management and the technological limitations making smart cities smart is the use of associated technolo-
of traffic light control devices, contribute to increasing congestion, gies with physical, electrical and communication technologies
environmental and noise pollution. The applications involving
the Internet of Things (IoT) concept have been growing around information (ICT) [9], [11], [12].
the world. Concerning the problems found in cities with large Congestion is one of the main agents responsible for urban
populations, the IoT applications aim technological solutions disorder. According to D. Schrank et al, in 2014 the congestion
that improve the population living conditions. Most traffic light caused the north americans about 6.9 billion hours of extra
controller systems are locally programmed. In this case, to control travel, consuming around 3.1 billion gallons of extra fuel,
all traffic lights in a city, besides to be a slow task, it is not
possible to determine the traffic light’s operating state. In some causing a traffic jam cost of 160 billion dollars [13].
cases a centralized control is used, allowing information exchange For current traffic lights systems to be able to manage the
with all traffic lights on the network. This paper seeks to study demand for future vehicles it is essential that it becomes a
and develop a traffic light controller electronic circuit, with smart device.
centralized control topology, able to communicate with other
The Internet of Vehicles (IoV), a scientific branch dedicated
wireless network traffic lights. All characteristics of the designed
circuit were evaluated by practical system control and monitoring to studies related to the information exchange between vehi-
tests. Finally, the electronic circuit proved to be viable in terms cles, through embedded sensor circuits, allow the collection of
of smart cities applications. road conditions [14], [15]. Therefore another utility of wireless
Index Terms—Smart traffic light, smart city, smart mobility. traffic lights is the possibility of integration with vehicles that
have IoV applications.
I. I NTRODUCTION
The present work, based on the IoT and ICT concepts,
Congestion, environmental pollution, noise pollution and intends to develop a traffic lights controller electronic circuit
increased health problems are some related problems to the capable of controlling and monitoring the state of each lamp,
badly planned city growth [1]–[5]. Although urban growth communicate with other network traffic lights, exchanging
seems to bring only problems, the constant technolog-ical system information over a wireless network, providing, in this
advances in the engineering area seek not only to reduce such way, an enhancement in the current traffic controller systems.
problems, but also offers new features that make it easier for Therefore, the following sections will describe the choice and
citizens, one of them is called Internet of Things (IoT). functioning of each block of the semaphore control electronic
The IoT term has several interpretations and definitions, ac- circuit.
cording to [6], the IoT term can be understood, in its physical
aspects, as being: “a network of items - each embedded with II. S MART T RAFFIC L IGHT
sensors - connected to the internet”.
In 2011, CISCO IBSG estimated that by 2020 there will be Although the vehicles flow control is done by the
about 50 billion devices connected to the internet [7]. In 2018, semaphore, the control mechanism can be done in different
the Qualcomm technology company assumed that at 2026 ways. According to [16], a controller may be set in two ways:
there will be about 6 billion connections of only applications • Fixed time: has constant semaphoric plane, that is, it does

involving IoT [8]. not change the time of its lights according to the flow
With the number of IoT applications growing increasingly variation;
and the constant need for strategic solutions that make life in • Acted: The determination of the green time based on the

large urban centers pleasant to citizens, there’s the emergence data by vehicle detector sensors in its approximations.
of the so-called smart cities. According to [9], [10], a city is Regarding the control mode, according to [16], traffic lights
classified as smart city when it is able to gather technology, may have:
government and society together to enable management with • Local control: the control is done directly in the controller
intelligent economy, mobility, environment, people, life and system, in loco;
governance. • Centralized control: the control is done by a single central

This study was financed in part by the Coordenação de Aperfeiçoamento device, controlling the network traffic lights operating
de Pessoal de Nı́vel Superior - Brasil (CAPES) - Finance Code 001. state.

978-1-7281-2949-5/19/$31.00 ©2019 IEEE 155

Authorized licensed use limited to: Universidad de Sevilla. Downloaded on May 08,2024 at 21:30:31 UTC from IEEE Xplore. Restrictions apply.
 2019 Sixth International Conference on Internet of Things: Systems, Management and Security (IOTSMS)

The main sensors types used to measure vehicles flow for was considered the insertion of a pedestrian button for each
a given intersection are: magnetic and infrared sensors and controller circuit. The electronic system high-level architecture
digital cameras [2], [17], [18]. diagram can be seen in Figure 2.
The wireless traffic lights used together with several wire-
less sensors, allows the creation of strategic plans in order POWER SOURCE
to control and synchronize the traffic and for prioritizing
certain routes and keeping constant, the vehicles speed, for RTC

example. Another major research area is related to Internet

ACTUATOR ACTUATOR
of Vehicles (IoV), a scientific branch dedicated to the studies AC DC
of communication between vehicles, that through embedded MICROCONTROLLER

circuits equipped with sensors, allow the road conditions

SENSOR SENSOR
collection [14], [15]. Another useful feature of wireless traffic AC DC
lights, therefore, is their ability to integrate with vehicles that
PEDESTRIAN MODULE
have IoV applications. BUTTON RF

Figure 1 illustrates the use of wireless traffic light.

Fig. 2. High-level architecture diagram of traffic light controller electronic
circuit.

The following sections will discuss each high-level archi-

tecture presented in Figure 2.
A. Electronic Power Supply Circuit
In order to provide adequate voltage levels for each circuit
part, the power supply circuit shown in Figure 3 was designed.
It is possible to check the varistor use, to protect the controller
circuit from over-voltage peaks.

LINE VDC1
J1 +5V
1 2 3
Vac_in1 Vdc_out
2
VAR1 C1

1 4
Vac_in2 GNDout
NEUTRAL
Fig. 1. Traffic lights comparison, wired (upper avenue) and wireless (lower GND
avenue) communication. U1
VOUT
GND
VIN
DC_LAMP
According to Figure 1, a centralized traffic light control J2
1
2
3

2
system communication using cable, only one controller per 1 C4
+5V GND
+3V3
intersection is required. However, when removing all commu- C2
nication cables, it is necessary that all semaphores from the GND

same intersection have an internal controller. Although there GND

is one controller per semaphore, the electronic circuit becomes

Fig. 3. Electronic system power supply circuit.
simpler.
When all traffic lights have wireless communication, its According to Figure 3, the circuit has a power supply that
physical installation becomes simpler, just having to place the converts the incoming AC voltage (127/220 V) to the DC
traffic light in the desired location and connect it to the electric voltage (5 V), voltage required by the Radio Frequency (RF)
grid. module. For the microcontroller, a voltage regulator module
III. E LECTRONIC C IRCUIT for 3.3 V was used. In order to control DC lamps, a 12 V
DC supply voltage input terminal has been inserted to provide
As a design requirement, we thought of an approach to a energy for the lamp, DC LAMP.
necessary minimum circuit for focal groups control (set of
lamps) from a traffic light, proposing not only traffic light B. Microcontroller Electronic Circuit
technological improvements, but also developing a technology Adjusting the lamp period, reading the current sensors and
that is a low cost solution. Two control and monitoring circuits the communication with the RF module are some of the
were designed, one for alternating current (AC) lamps and microcontroller circuit performed tasks, according to Figure
another one for direct current (DC) lamps, in order to control 4. For that, STMicroelectronics STM32F103C8T6 microcon-
the different types of lamps used at traffic lights. A burnt troller with Cortex-M3 32-bit ARM core was used.
light detection circuit was also developed for each lamp type. This microcontroller has an internal Real Time Clock (RTC)
In order to implement the pedestrians priority detection, it peripheral to the chip, requiring the use of only one external

156
Authorized licensed use limited to: Universidad de Sevilla. Downloaded on May 08,2024 at 21:30:31 UTC from IEEE Xplore. Restrictions apply.
 2019 Sixth International Conference on Internet of Things: Systems, Management and Security (IOTSMS)

U2 U3
TABLE II

RX_STM32
SENS_AC_R_S 10 18
PA0-WKUP PB0
SENS_AC_Y_S 11 19
PA1 PB1 ACTUATION TRUTH TABLE , WITH THE DC LAMP BURNT.
RX_MSP
RX_MSP
TX_MSP
TX_MSP
RST_RF
RST_RF
12 20
PA2 PB2/BOOT1
13 39 RST_RF

GND
GND
3.3V
3.3V
PA3 PB3/JTDO
SENS_AC_G_S 14 40 SENS_DC_R_S
5V
5V

PA4 PB4/JNTRST
SENS_AC_R_P 15 41 SENS_DC_Y_S R1
SENS_AC_G_P 16
PA5
PA6
PB5
PB6
42 TX_STM32 Input MOSFET BJT Output Situation
1
7
2
8
3
9
4
10
5
11
6
12
17 43
PA7 PB7
0V Blocking Blocking 3.3 V Lamp burned
AC_R_S 29 45 SENS_DC_G_S R2
PA8 PB8
3.3 V Driving Blocking 3.3 V Lamp supposedly on
RX_STM32

TX_STM32

RST_RF

+5V +3V3 AC_Y_S 30 46 SENS_DC_R_P

PA9 PB9
AC_G_S 31 21 SENS_DC_G_P
PA10 PB10
AC_R_P 32 22 DC_R_S
PA11 PB11
GND AC_G_P 33 25 DC_Y_S
PA12 PB12
SWDIO 34 26 DC_G_S GND
PA13/JTMS/SWDIO PB13
C3 SWDCLK 37 27 DC_R_P
PA14/JTCK/SWCLK PB14
P2 +3V3 38 28 DC_G_P
GND PA15/JTDI PB15
1
Considering the table and the circuit, it is noted that the
2

SWDIO 5 2 C5
2 Y1 OSC_IN/PD0 PC13-TAMPER-RTC
SWDCLK 6 3
3 OSC_OUT/PD1 PC14-OSC32_IN GND
4
4 C6 PC15-OSC32_OUT damaged lamp can be detected when both the MOSFET (Q1)

2
1

GND Y2
44
BOOT0
GND C7 and the BJT (Q16) are blocked.

1
7
NRST GND
BT1
N P 1
GND VBAT
R3 C8 24 23
VDD_1 VSS_1
+3V3
36
48
VDD_2
VDD_3
VSS_2
VSS_3
35
47
D. Electronic AC Current Sensor Circuit
GND GND 9 8
VDDA VSSA
C9 C10 C11 C12 C13 C14 To control and monitor AC lamps, the electronic circuit
GND
of Figure 6 was designed. It has the same idea of burned
GND GND GND GND GND GND
lamps detection seen in the previous section, but because its
circuit has a AC power supply, the circuit has undergone some
Fig. 4. Microcontroller schematic circuit.
adjustments.

R29
precision crystal, Y2. The RTC peripheral will be used as a LINE 2
J8

1
clock source for the lamp timers adjustment as well as for the

4
Q11 R31
U4
synchronization between the network semaphores. U6
+3V3

1 6
D1

SENS_AC_R_S
R33
C. Electronic DC Current Sensor Circuit R35
NEUTRAL 2 5

2
+5V
One way to detect the light bulb state, whether it is working R37 R38 Q13
4
GND
normally or if it is burned, is to measure a current signal. For R41
C15
AC_R_S
the particular case of a DC lamp, the current sensing circuit
GND GND GND
of Figure 5 was designed. It also allows to control the focus
groups via the MOSFET transistor (Q1).
Fig. 6. Schematic burnt lamps sensor circuit, with AC power.

DC_LAMP J3 DC_LAMP
In this way, the lamp switching signal is sent to the
SENS_DC_R_S

2
1
R31
circuit input, AC R S, referring to the lamp AC switching
DC_R_S

that represents the red signal. The lamp status is read via
R35
Q1 R33 the circuit output, SENS AC R S. The loads are driven by
Q16
R6
R4
an optocoupler circuit (U4) which provides the trigger signal
for TRIAC, which, when in conduction, provides the supply
signal for the AC load. To check if the load is conducting
GND GND GND GND GND
current, another optocoupler circuit is inserted and when the
lamp is working correctly, emits a 0 V output signal. The
Fig. 5. Schematic burnt lamps sensor circuit, with DC power. circuit behavior of Figure 6 can be analyzed by Tables III and
IV.
Observing the input signal (DC R S - controlling the red
lamp signal) behavior, the voltage levels of their respective
TABLE III
output (SENS DC R S) can be found. Tables I and II sum- ACTUATION TRUTH TABLE , WITH THE AC LAMP WORKING NORMALLY.
marize the circuit function.
Input TRIAC Output Situation
TABLE I 0V Blocking 3.3 V Lamp off
ACTUATION TRUTH TABLE , WITH THE DC LAMP WORKING NORMALLY. 3.3 V Driving 0V Lamp on

Input MOSFET BJT Output Situation

0V Blocking Driving 0V Lamp off
3.3 V Driving Blocking 3.3 V Lamp on TABLE IV
ACTUATION TRUTH TABLE , WITH THE AC LAMP BURNT.

In order not to damage the microcontroller’s input pins, in Input TRIAC Output Situation
the output SENS DC R S, the resistors R31 and R33 values 0V Blocking 0V Lamp burned
were adjusted so that the signal did not exceed the 3.3 V input 3.3 V Driving 0V Lamp supposedly on
voltage limit.

157
Authorized licensed use limited to: Universidad de Sevilla. Downloaded on May 08,2024 at 21:30:31 UTC from IEEE Xplore. Restrictions apply.
 2019 Sixth International Conference on Internet of Things: Systems, Management and Security (IOTSMS)

E. XMesh Network and RF Module T – TW/TW

PM1 PM2
Several wireless technologies operate under different cir-
cumstances. Similar works involving wireless semaphores VM4 3
VM3
communication make use of the ZigBee communication pro-
1 VM1
tocol creating a mesh topology network, [17], [19]–[21]. VM2

Although such protocol defines a data packet delivery algo- 2

rithm, it does not implement a complete routing table. The PM3
VM5
implementation of an optimized data packet delivery algorithm
is essential for IoT applications, where most of the devices
have low processing power. A developing technology from a
doctorate research at Unicamp [22], whose route algorithm is STAGE 1 STAGE2 STAGE 3

performed through mathematical modeling, allows its use in

microcontrolling and embedded systems. The network block
diagram, titled XMesh, can be seen in Figure 7.

FULL CYCLE
Device

M4IoT
Fig. 9. Model of an T intersection TW/TW.

N4IoT

focal group (with two lamps), both 127/220 V AC and 12 V

Hardware
DC, in addition it is able to perform a pedestrian button action.
XMesh
G. Microcontrollers Firmware
After developing the electronic traffic light controller cir-
Fig. 7. XMesh network model.
cuit, the microcontroller firmware was programmed. Due to
The N4IoT block refers to the network layer that has the wireless communication, several control, monitoring and
mathematical modeling for routes, allowing the addition of security routines were elaborated in order to guarantee the
numerous devices to the network. The M4IoT block refers to identification of all network traffic lights, as well as the
the protocol layer, encrypted by means of elliptic curves. It is robustness to eventual failures.
able to create a Virtual Private Network (VPN) for each IoT When powering up the controller system, it runs a config-
application, called Virtual Private Device Network (VPDN). uration routine to determine its operating mode. This routine
This feature allows a wireless traffic light network to operate performs message exchange with the network gateway device.
simultaneously with a wireless sensors network. In this way, At the end of the configuration routine, the semaphore is able
the present work was chosen to use the XMesh network. to be controlled and / or monitored remotely.
To detect the traffic lights fault at an intersection, either
F. Traffic Controller Circuit due to electrical problems or communication failure, several
After analyzing all circuits described previously, they were safety routines have been implemented. In the event of any
all assembled on the same Printed Circuit Board (PCB) and system failure, immediately all intersection’s running traffic
the final result can be seen in Figure 8. lights switch to the traffic light fault state (flashing yellow)
and send a message to the network gateway device.
Finally, for each intersection type addressed, a state machine
was developed capable of performing its correct operation.
Therefore, when changing the intersection type, all intersection
traffic lights adapt to the new mode of operation.
IV. R ESULTS
For the traffic light controller to be able to operate in
real conditions at a traffic intersection, it was necessary to
search the main types of intersections. According to [23],
the intersections that frequently use semaphores have three
(T-shaped) and four (cross-shaped) approximations. Figure 9
Fig. 8. Traffic Light Controller Circuit. exemplifies the two-way T intersection (TW / TW) operation.
A traffic light complete cycle consists of a stages sequence,
The PCB presented in Figure 8, allows to control and according to Figure 9. One way to display the all stages
monitor a vehicle focal group (with three lamps), a pedestrian execution of all semaphores in an intersection is through a

158
Authorized licensed use limited to: Universidad de Sevilla. Downloaded on May 08,2024 at 21:30:31 UTC from IEEE Xplore. Restrictions apply.
 2019 Sixth International Conference on Internet of Things: Systems, Management and Security (IOTSMS)

VFG1
As described in Subsection III-B, the adopted microcon-
trolling system performed the execution of all the necessary
PFG1 routines to establish a wireless traffic management of all
semaphores in the network. The XMesh network proved to be
VFG2 feasible to manage the manage the data packets through the
N4IoT network layer, in addition to administering the message
PFG2
exchange through the M4IoT protocol.
VFG3
The actuator and sensor AC/DC circuits described in Sub-
sections III-C and III-D demonstrated their versatility in
PFG3 working with focal groups with 127/220 V AC and 12 V DC
power.
t The results obtained in Section IV for the T intersection
TW/TW demonstrated the designed traffic light controller cir-
Fig. 10. Temporal diagram of an T intersection TW/TW, theoretical.
cuit functionality, allowing all traffic lights at the intersection
to operate in synchronism.
Data collection by IoV applications, such as traffic accident
detection, shared with the proposed system, enables the en-
hancement of traffic light management systems, blocking and
creating alternative routes to not only avoid the traffic jams,
but also avoid new accidents.
Therefore, the designed traffic light controller circuit, be-
sides offering a new alternative to today’s traffic light control
systems, allows the study and development of technologies
that contribute to the transformation of large urban centers
into smart cities.

ACKNOWLEDGMENT
This study was financed in part by the Coordenação
de Aperfeiçoamento de Pessoal de Nı́vel Superior - Brasil
(CAPES) - Finance Code 001.

Fig. 11. Temporal diagram of an T intersection TW/TW, practical. R EFERENCES

[1] B. Boriboonsomsin and M. Barth, “Real-World CO 2 Impacts of Traffic
Congestion,” Transportation Research Record, no. 951, pp. 1–23, 2008.
time diagram, as shown in Figure 10. At the operation cycle [2] L. Qi, M. Zhou, and W. Luan, “Emergency traffic-light control system
design for intersections subject to accidents,” IEEE Transactions on
end, the same time diagram is started again, creating a traffic Intelligent Transportation Systems, vol. 17, no. 1, pp. 170–183, 2016.
management plan. [3] S. Agarwal and B. Swami, “Road traffic noise, annoyance and
To perform the controller circuit test operating at the con- community health survey - A case study for an Indian city,” Noise and
Health, vol. 13, no. 53, pp. 272–276, 2011. [Online]. Avail-
ditions depicted in Figure 9, the experiment was carried out able: http://www.noiseandhealth.org/article.asp?issn=1463-1741;year=
with the assembly of three equal PCBs, one of them shown 2011;volume=13;issue=53;spage=272;epage=276;aulast=Agarwal;t=6
in Figure 8. [4] C. Li and S. Shimamoto, “An open traffic light control model for reduc-
ing vehicles’CO2 emissions based on etc vehicles,” IEEE Transactions
In order to visualize the signals sent to each VFG and PFG on Vehicular Technology, vol. 61, no. 1, pp. 97–110, Jan 2012.
of each intersection, a logic analyzer was used. The traffic [5] Y. Huang, Y. Weng, and M. Zhou, “Design of traffic safety control
systems for emergency vehicle preemption using timed petri nets,” IEEE
lights data can be seen in Figure 11. In this test, we considered Transactions on Intelligent Transportation Systems, vol. 16, no. 4, pp.
a green signal having 10 s and 2 s for the yellow signal. 2113–2120, Aug 2015.
According to Figure 11, the traffic light control circuits [6] The institute. (2014) Special report: The internet of things.
[Online]. Available: http://theinstitute.ieee.org/static/special-report-the-
behave as expected. The only difference is that during the internet-of-things
tests it was decided to implement the blinking red routine for [7] D. Evans, “The Internet of Things How the Next Evolution of the
pedestrians, indicating that the vehicles traffic light will turn Internet Is Changing Everything,” Cisco Internet Business Solutions
Group, no. April, 2011.
green, according to the brazilian manual of traffic signaling [8] Qualcomm, “Leading the lte iot evolution to connect the
[24]. massive internet of things,” qualcomm.com, 2018. [Online].
Available: https://www.qualcomm.com/media/documents/files/leading-
the-lte-iot-evolution-to-connect-the-massive-internet-of-things.pdf
V. C ONCLUSION [9] S. P. Mohanty, U. Choppali, and E. Kougianos, “Everything you wanted
to know about smart cities: The Internet of things is the backbone,”
The system power supply circuits proved capable of pro- IEEE Consumer Electronics Magazine, vol. 5, no. August, pp. 60–70,
viding adequate voltage levels for each designed PCB part. 2016.

159
Authorized licensed use limited to: Universidad de Sevilla. Downloaded on May 08,2024 at 21:30:31 UTC from IEEE Xplore. Restrictions apply.
 2019 Sixth International Conference on Internet of Things: Systems, Management and Security (IOTSMS)

[10] IEEE Smart Cities Community. (2018) Leading the

lte iot evolution to connect the massive internet of
things. [Online]. Available: https://www.ieee.org/membership-
catalog/productdetail/showProductDetailPage.html?product=
CMYSC764\&N=4294925450+4294925418\&tcType=General
[11] S. P. Mohanty, “Smart Technologies: The Key for Sustainable Smart
Cities,” IEEE Consumer Electronics Magazine, vol. 7, no. march, pp.
3–4, 2018.
[12] S. Gautama, “Smart Mobility,” 60th International Symposium ELMAR,
no. September, pp. 16–19, 2018.
[13] D. Schrank, B. Eisele, T. Lomax, and J. Bak, “Urban mobility,”
Urban Mobility Scorecard, no. August, pp. 1–39, 2015. [Online].
Available: http://d2dtl5nnlpfr0r.cloudfront.net/tti.tamu.edu/documents/
mobility-scorecard-2015-wappx.pdf
[14] L. Chen and C. Chang, “Cooperative traffic control with green wave
coordination for multiple intersections based on the internet of vehicles,”
IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 47,
no. 7, pp. 1321–1335, July 2017.
[15] X. Wang, Z. Ning, X. Hu, L. Wang, B. Hu, J. Cheng, and V. C. M.
Leung, “Optimizing content dissemination for real-time traffic manage-
ment in large-scale internet of vehicle systems,” IEEE Transactions on
Vehicular Technology, vol. 68, no. 2, pp. 1093–1105, Feb 2019.
[16] Conselho Nacional de Trânsito, Sinalização Semafórica. Manual
Brasileiro de Sinalização de Trânsito, 2014.
[17] Q. Wang, J. Zheng, H. Xu, B. Xu, and R. Chen, “Roadside Magnetic
Sensor System for Vehicle Detection in Urban Environments,” IEEE
Transactions on Intelligent Transportation Systems, vol. 19, no. 5, pp.
1365–1374, 2018.
[18] L. P. J. Rani, M. K. Kumar, K. S. Naresh, and S. Vignesh, “Dynamic
traffic management system using infrared (IR) and Internet of Things
(IoT),” 3rd IEEE International Conference on Science Technology,
Engineering and Management, pp. 353–357, 2018.
[19] J. Pedro, “Sistema de Comunicação Wireless para Controlo e Gestão de
Tráfego,” Master’s thesis, Instituto Superior Técnico de Lisboa, Portugal,
2016.
[20] G. P. B. Silva, “Traffic Bee: Proposta de Aplicação da Tecnologia ZigBee
em Sistema Inteligente de Semáforos,” Master’s thesis, Instituto Federal
do Rio Grande do Norte, Brasil, 2018.
[21] R. Sundar, S. Hebbar, and V. Golla, “Implementing Intelligent Traffic
Control System for Congestion Control , Ambulance Clearance , and
Stolen Vehicle Detection,” IEEE Sensors Journal, vol. 15, no. 2, pp.
1109–1113, 2015.
[22] P. D. G. LUZ, L. T. Manera, and M. M. Donatti, “Dispositivo de mon-
itoramento da rede pública de iluminação, sistema de monitoramento e
gerenciamento da rede pública de iluminação e método de comunicação
sem fio,” Brasil, Número do Processo: BR 10 2017 023 482 7, 2017.
[23] J. F. C. Martins, “Seleção de Interseções com Potencial de Redução
da Sinistralidade - Aplicação do HSM,” Master’s thesis, Faculdade de
Engenharia da Universidade do Porto, Portugal, 2013.
[24] Conselho Nacional de Trânsito, Sinalização Semafórica. Manual
Brasileiro de Sinalização de Trânsito, 2014.

160
Authorized licensed use limited to: Universidad de Sevilla. Downloaded on May 08,2024 at 21:30:31 UTC from IEEE Xplore. Restrictions apply.