Traffic Profiling and Intelligent Street Lights for a Smart City

Project Title

Traffic Profiling and Intelligent Street Lights for a Smart City

Project Area of Specialization Artificial IntelligenceProject Summary

In developing nations, many expanding cities are facing challenges that result from the overwhelming numbers of people and vehicles causes’ severe traffic congestion. Collecting real-time, reliable and precise traffic flow information is crucial for urban traffic management. The main purpose of this project is to do traffic profiling in the cities with the help of counting and classification of vehicles on the road in real-time. We are using computer vision techniques for counting and classification of vehicles based on deep learning algorithms, after counting and classifying this all data will be send it  to cloud for data analysis and modeling.at this smart centre we will take different predictions regarding traffic congustion of specific day and time with the help of machine learning.This will help us to minimize the traffic congestion. we are implementing all the algorithms on embedded devices, this is an strenuous task to implement neural networks on limited computational power devices, in this regards the selection of neural network algorithm is very important.

Secord major problem in the world is power crises  World Economic Forum's Global Shapers Survey believe power crises is the most serious issue affecting the world today. about 25%  of the power is consumed by street lights in the world, so we proposed a solution for this to make the street lights intelligent. Intelligent street lighting refers to public street lightning that adapts to movement by pedestrians, cyclists and cars. Intelligent street lighting also referred to as adaptive street lighting, dims when no activity is detected, but brightens when movement is detected. This type of lighting is different from traditional, stationary illumination, or dimmable street lighting that dims at pre-determined times. in this we will actually control the intensity of the light, whenever the vehicle is detected the intensity of the light goes higher.

Project Objectives

• to minimize the traffic congestion
• to improve the utilization of human resource(in this we will improve the utilization of traffic warden by predicting the traffic congestions at different places,if our system predict that there will be a traffic congestion at kashmir highway at 1600 then this will also tell the numbers of traffic wardens required at that point)
• to minimize the power utilization by street lights(by controlling the intensity of street lights)

Project Implementation Method

• we are implementing this idea on a real scenrio by using state of the art Artificial intelligance development kit NVIDIA jetson nano,that is specially designed for the neural networks,in this we are using quantized sigle shot detecter mobilenet for classification of vehicle and other computer vision algotithms for thr reliable counting of vehicles,we are using camera with NVIDIA Jetson nano for counting and classification of vehicles.then we send all this data to the cloud and do data analysis and modeling we are using schtochistics models for graphically represenation of traffic condition and we use machine learning algorithms for prediction of congestion.we have check our project on kashmir highway and peshawar road on real time and the results were amazing.
• 2nd task is to control the intensity of the LED street lights,we speciallly designend our own circuit that can control the intesity of street light.

Benefits of the Project

• as we know the number of vehicles on the roads of pakistan are increasing day by day which causes severe congestion on peak hours and many times the ambulances,school,office vans, and VIP's got stuck in traffic we are not able to invest huge amount for the wide roads day by day,this can only be controll by using smart technoligies,so our project is the solution of this probrlem.
• we are using state of the art embedded devices and alogorithms in our projectso this will prove to be fruitful for us in the professional life.
• we are using tools like NVIDIA,linux,python,open cv,deep learning algorithms(YOLOV3,SSD MOBILENET),algorithms of computer vision,circuit analysis and design tools.these all tools can help us in future.
• this product that can be able to predict the traffic condiition will make good place in market.
• this is cost effective approach for traffic profiling.
• this will solve the traffic issue and energy crises issues.
• the data that collected from roads can be used for many other purposes,because at this time in our pakistan we dont have data sets for our enviornment this is the main reason that no one can implement AI on the enviornmoment,our project will help in this regards to collect the data from roads and anyone can implement some other system by taking differnet predictions.
• the intesity controling circuit will help us to minimize the power cunsumption in cities,the intesity cotrol lights are also enviornment friendly.

Technical Details of Final Deliverable

• we will design the device that can be able to count and classify the vehicles vehicles(contain camera and NVIDIA jetson nano mainly) this will place on the roads and and these devices will count and classify vehicles.
• after all the data analysis and modeling we will come up with the system that can predict the traffic congestion and traffic conditions at different roads of the city.
• we designed circuit that control the intesity of the LED street light,when ever the vehicle is detected the intensity of the next 3,4 lights will goes to 80% and when there is no vehicle on the road the the light glow with 20% intensity.

Final Deliverable of the Project HW/SW integrated systemCore Industry TransportationOther Industries Energy Core Technology Artificial Intelligence(AI)Other Technologies Internet of Things (IoT), Cloud Infrastructure, Big DataSustainable Development Goals Industry, Innovation and Infrastructure, Sustainable Cities and CommunitiesRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	41500
nvidia jetson nano board	Equipment	1	24000	24000
webcam camera logitech	Equipment	1	3000	3000
FAN and acrylic case for nvidia and camera	Equipment	1	2000	2000
electronics componets and pcb board for intensity control circuit	Equipment	1	3000	3000
LED street light for testing	Equipment	1	6000	6000
led screen that is conected with nvidia board	Equipment	1	3000	3000
pcb printing and other reports printings	Miscellaneous	1	500	500