Travel time prediction for dynamic routing using Ant Based Control

Currently most car drivers use static routing devices based on the shortest distance between start and end position. But the shortest route can differ from the shortest route in time. To compute alternative routes it is necessary to have good prediction models of expected congestions and a fast algorithm to compute the shortest path while being able to react to dynamic changes in the network caused by special incidents. In this paper we present a dynamic routing system based on Ant Based Control (ABC). Starting from historical traffic data, ants are used to compute and predict the travel times along the road segments. They are finding the fastest routes not only looking to the past and present traffic conditions but also trying to anticipate and avoid future congestions.