Red Swarm: Reducing travel times in smart cities by using bio-inspired algorithms

Abstract

This article presents an innovative approach to solve one of the most relevant problems related to smart mobility: the reduction of vehicles’ travel time. Our original approach, called Red Swarm, suggests a potentially customized route to each vehicle by using several spots located at traffic lights in order to avoid traffic jams by using V2I communications. That is quite different from other existing proposals, as it deals with real maps and actual streets, as well as several road traffic distributions. We propose an evolutionary algorithm (later efficiently parallelized) to optimize our case studies which have been imported from OpenStreetMap into SUMO as they belong to a real city. We have also developed a Rerouting Algorithm which accesses the configuration of the Red Swarm and communicates the route chosen to vehicles, using the spots (via WiFi link). Moreover, we have developed three competing algorithms in order to compare their results to those of Red Swarm and have observed that Red Swarm not only achieved the best results, but also outperformed the experts’ solutions in a total of 60 scenarios tested, with up to 19% shorter travel times.