Celer: A Smart Fleet Management System (Optimizing Traffic Flow in New York City)

Abstract

As society moves closer to fully autonomous vehicles, it must eventually make vehicles work together. This would reduce traffic jams, reduce cost of trips, reduce overall travel time, reduce the environmental impact, and reduce the number of casualties to traffic. [1] However, society's focus has mostly gone towards making the vehicles autonomous and not towards making a system that would manage a set of robo-taxis. This gap in research should be thoroughly explored because although autonomous vehicles are safer, they are not necessarily more efficient in reducing traffic jams and the cost of trips. [6] There have been many promising studies in tackling individual issues that such a system would face. These include finding an efficient route from point A to point B [2, 3], optimizing intersections [4], tackling road hazards [6], and more. By combining many preexisting algorithms into one system, Celer attempts to optimize traffic flow in New York City and explore the problem of car interconnectivity. Celer is able to reconstruct a map of New York City and uses taxi data from 2015 to simulate real world conditions. Overall, Celer improved trip time and profits substantially and showed a promising solution to the fleet management problem.