Smart Traffic Light
Project overview
As part of the SmaRTF hackathon, our team "Code & Klystron" developed a Proof of Concept for a Smart Traffic Light. The device is portable, has embedded computer vision, Bluetooth connectivity for settings changes, RFID key security measures, and a LoRaWAN protocol antenna for wireless communication between traffic lights.
Moment of winning the SmaRTF Hackathon
The project won 1st place in the first embedded hackathon competition involved all of Ukraine. Our work was presented in Ukrainian news.
Sources:
ICTV news on our victory and hackathon: https://youtu.be/0pkym2vUayA?t=375
University (organizer) article on hackathon: https://kpi.ua/2021-12-SmaRTF
Business Value
Traffic congestion is a major issue in many cities around the world. Congestion leads to longer travel times, wasted fuel, increased emissions, and driver frustration. Traditional traffic lights operate on fixed timing schedules and are not adaptive to real-time traffic conditions, resulting in traffic backups and delays.
The Smart Traffic Light device offers a solution to these problems. By using computer vision technology, the traffic light can analyze traffic patterns in real-time and adjust the timing of the traffic signals to optimize traffic flow. Additionally, the device's portability makes it a valuable asset during road repair works, where traditional traffic light installations may not be feasible.
The Bluetooth connection allows workers to remotely change timing settings, reducing the need for manual adjustments. The use of RFID technology provides a secure method for accessing and modifying the traffic light's settings.
Technical details
Project design had several stages. In the first stage we developed general idea and came-up with possible solutions. Right after that we applied Risk Analysis methodology DREAD-STRIDE to assess possible threats and find mitigation strategy.
STRIDE methodology
From STRIDE point of view we identified two main risks in the project: DoS (physical harm or hacking) and spoofing. We aimed on preventing some of risk sources on the design level. To prevent DoS and Spoofing we added special level of authorization with RFID. Therefore, hacker can no longer spam with requests as they are ignored without RFID card. Also, Spoofing is mitigated with Diffie-hellman key exchange encryption technique.
Talking hardware-wide, we used an ESP-32 microcontroller with an embedded camera, as opposed to an expensive microcomputer like Raspberry Pi. This cost-efficient solution provides the necessary computational power for the device's functions. The device's computer vision algorithms analyze real-time traffic data and adjust signal timing accordingly.
The device also features Bluetooth connectivity, allowing for real-time configuration changes, while RFID key technology ensures only authorized personnel can access the settings.
For wireless communication between traffic lights, we utilized a LoRaWAN protocol antenna, which provides long-range and low-power communication. The synchronized lights help to ensure efficient traffic management and reduce congestion.
Model of the device core
In this 3D model of the device, you can see PCB implemented all circuits shown above. It has an input for the LoraWAN antenna and special high-power input for green, yellow and red LEDs. The black device is a core brain and eyes, ESP-32, with an embedded camera. And the white one is an additional controller, Arduino Pro Mini, to work with sensors and reduce the pin load from ESP-32.
Results
Our team won first place in the SmaRTF hackathon with our Smart Traffic Light proof-of-concept device. The device demonstrated that computer vision technology can be used to optimize traffic flow and reduce congestion. The device's portability, wireless communication capabilities, and secure access settings make it a valuable asset for traffic management.
Future Works
There are several potential directions for further development of the Smart Traffic Light device. One possible avenue is the integration of machine learning algorithms to improve the device's computer vision capabilities. Another is the use of data analytics to optimize traffic flow and reduce congestion further. The device could also be modified to integrate with existing traffic management systems, providing real-time data for improved decision-making. Overall, the Smart Traffic Light device has significant potential for future development and implementation in smart cities.
Acknowledgements
I can not help telling you about the best team I ever worked with! Drop any team members, and I doubt we achieve the same results. The whole team worked remotely and never met all together. Everybody had their separate field of work - one works on the mechanical side; another does the hardware part from modelling to assembling, and the next add Computer Vision to the model. This teamwork experience showed that people could work together as precisely as an atom clock, which is the merit of each of you: Kostya Sirov, Andrey Ivanchenko, Demid Strukov and mine.
