AMS BalanCity

Smart Traffic Load Balancing for Sustainable Cities

Managing motorized traffic flows is an important issue in every major metropolitan area. Especially in Amsterdam, which should take into account economic productivity loss due to traffic congestion but also noise and air pollution. We propose a route planning system which considers multiple, collective and concurrent objectives concerning congestion and pollution on various roads, and balances traffic over the road network in a highly effective way. The project will produce a technological proof-of-concept, including a demonstrator website, of such a load balancing system, using techniques from artificial intelligence, traffic science, and multi-agent systems, as well as open map, traffic, and pollution data.

Problem statement
Managing motorized traffic flows is an important issue in every major metropolitan area. This is true in particular for Amsterdam, due to its high population density, congestion, and noise and air pollution problems. Traffic management aims to reduce these problems, but it cannot normally influence routes chosen by individual vehicles. On the other hand, many modern individual route planning and navigation systems take into account real-time congestion information, but they do not take into account collective interests such as limited capacity or pollution constraints of alternative routes. Especially when more and more individuals use such individual route planning and navigation systems this can lead to situations of large numbers of vehicles simultaneously choosing the ‘fastest’ alternative route through limited capacity, densely populated urban areas when a motorway is congested – thus polluting and congesting roads where this does most harm and, ironically, possibly even slower routes for the individual users.

Approach
We propose an approach where individual route planning and navigation systems on the one hand and collective considerations on the other are not completely independent and sometimes counteractive processes. In contrast, route planning and navigation systems should be able to take into account multiple, collective objectives concerning congestion and pollution on various roads in the complete road network, and help in balancing traffic over the network in a highly effective way. This means that a heavy goods diesel truck may automatically be assigned a route avoiding a certain road in a densely populated area which already receives a large amount of traffic, even though its quickest route would use that road; while an electric vehicle would usually be assigned the quickest.

Objective
The project aims to develop and demonstrate a first version of a route planning system that takes into account multiple, collective objectives concerning congestion and pollution on various roads in the complete road network of Amsterdam, and that balances (distributes) traffic over the network in a principled way based on those objectives.

Project duration: December 2015 – September 2016

Partners: Delft University of Technology, Cygnify BV

Project leader:
Dr. M.T.J. Spaan, Assistant Professor, Delft University of Technology,
Phone: +31 15 278 1102
Mail: m.t.j.spaan@tudelft.nl

Related information:
Algorithmics at TU Delft
Cygnify

Contact:
Dr. M.T.J. Spaan, Assistant Professor, Delft University of Technology
Phone: +31 15 278 1102
Mail: m.t.j.spaan@tudelft.nl

Research output
Website

This project is an AMS Stimulus Project. The aim of Stimulus Projects is to give to new and existing AMS partners support to innovative research that has a strong upscaling potential. The projects should realize short-term research output, which act as a catalyst of a new solution direction, concept or approach.