Altering Traffic Light Timings to Reduce Car Usage

Abstract

Widespread car usage is a problem that cities worldwide are trying to control. They try to promote other forms of mobilities such as walking, cycling, and using public transport, while punishing car usage by forcing motorized traffic onto longer routes and pricing congestion. With reducing car usage as a goal without clogging up the road, we designed a genetic algorithm that was able to change behavior of agents based on the duration of traffic light timings. By elongating the length of green lights for non-motorized traffic (cyclists and pedestrians) compared to traffic lights for motorized traffic (cars and busses) with 35%, the model was able to improve the percentage of agents using eco-friendly transportation modes with 12.8 percentage points.