Two researchers at MIT have developed a model designed to help traffic flow more smoothly, the longer-term aim being to do away with road congestion and avoid unnecessary greenhouse gas emissions.
Traffic jams are not just a frustrating experience for car drivers; they also increase the volume of greenhouse gases emitted into the atmosphere. Two researchers at MIT, Carolina Osorio and Kanchana Nanduri, who have been trying to solve this dual problem, have now developed an algorithm to enable ‘smart’ control of a city’s traffic light systems. The software makes direct simulations of different situations, taking into account a set of parameters including driver behaviour, vehicle type, how complex an intersection is and so on. The ultimate aim is to monitor and synchronise in a ‘smart’ way – i.e. according to actual traffic flows around the city at any given time – all the traffic lights at all of the city’s intersections.
Theirs is a highly sophisticated approach, the aim being to optimise journey time between and along the main arteries, and so reduce overall petrol consumption – two aspects which current models of urban traffic flow management have so far not managed to achieve. Smart traffic lights have in fact been around for quite some time, but these systems are not sophisticated enough to take into account all the complex interactions of road traffic, such as the type of car and variations in driver behaviour. Moreover, they cannot predict how much a small alteration in daily traffic flow – such as changes in vehicle models and size, leading to lower petrol consumption, etc – will affect the amounts of greenhouse gases emitted. The new method purports to produce a realistic, but at the same time ‘tractable’, model.
Traffic jams: soon a thing of the past?
In order to create their urban traffic flow management model 2.0, the MIT researchers carried out simulations in the city of Lausanne in Switzerland. The work entailed modelling 17 key intersections and 12,000 vehicles per day in the city. The Osorio-Nanduri model pushes back the boundaries of what has so far been possible, due to a methodology which has deliberately reduced the amount of detail necessary, while retaining sufficient parameters to make accurate predictions and recommendations. Their data analysis method interprets complex detailed information at any moment. In the long term, this software could be used to do a lot more than smooth traffic flows by adapting traffic lights to the traffic flow at any given time. Osorio says that, going forward, this type of traffic simulation could also be used to optimise such planning decisions as choosing the best locations for car- or bike-sharing centres. Longer term, this ‘computer-friendly’ software could actually help urban management bodies to draw up city plans designed to generate more efficient traffic flows on the roads and streets, making environmental considerations a central facet of the system.