Intelligent Traffic Signal Control

Addressing the most challenging Bangalore Traffic which is rated to be 2nd most congested city in South Asia.


Team:

Sri Harsha P.

Sowmiya Nagarajan


My Role:

Product Designer


Deliverables:

A UX mockup for both the stakeholders involved in the project through navigable Figma prototype.


Time Period:

November 2018 [1 Months - Part time]

@ Bengaluru Tech Summit Global Hackathon 2018

 

Motivation & Proposed Solution

The traffic situation in Bengaluru is inevitably a matter of trouble for its commuters. A city with such a thriving population needs state-of-the-art Technology to attend to its high infrastructure needs. This project "AI for Traffic Signal Control" aims to maximize the tech utilization to come up with a frugal and adaptive solution for the problem.


Preliminary Objectives:

  1. 40% reduction in average waiting time

  2. Pre-emption system to ease movement of priority vehicles (Ambulances, Fire engines)

The proposed system aims to minimize waiting times for motorized vehicles in traffic junctions. The system extracts data like Queue Lengths and Average Speeds of vehicles from the live video feed. A Deep Q-Learning reinforcement model optimizes traffic light configuration based on extracted data.

An additional pre-emption system is planned to ease priority vehicle movement. Crisis regions are mapped for all junctions across the city. Location data of priority vehicles is used to activate green lanes for junctions.


High-level Architecture