Market analysis techniques from economic theory could help to improve the efficiency of urban traffic flows, underpinned by the use of vehicle-installed software interacting with smart traffic signal systems.

Market Analysis Approach to Optimising Urban Traffic Flow

Now that Google’s driverless car is looking as if it might really catch on, the concept of more efficient, entirely automated driving seems increasingly likely. This at least is the impression one gets when reading a paper published by Matteo Vasirani and Sascha Ossowski of the Rey Juan Carlos University in Madrid. These two academics have studied the way driverless cars might be used to optimise urban traffic flows so as to reduce traffic jams and cut down on the overall time spent on public road transport. Starting out from a classic market analysis of the supply and demand for available road space, especially at junctions, they looked at ways of improving the way city traffic signal systems work.

Software linking car to signal system

The Madrid-based academics looked into applying various traditional economic theories, focusing especially on an auction-based approach which offers a longer or shorter journey time depending on the value the driver ascribes to his/her own time. To optimise their concept, they pre-suppose the widespread use of intelligent autonomous vehicles, operated by software on behalf of their human owners, which interact with the traffic light infrastructure in order to negotiate safe and efficient passage through the road network. Under their concept, traffic signals would also be able to communicate with each other and would communicate signal phase and timing information to a vehicle’s system, which would in turn warn the traffic lights that it is approaching. The lights could then judge whether a time slot is available and assign it to the car. For example, when there is little traffic on the road, the timing of the lights, the space allotted to each vehicle and the number of cars passing through could all be modified in real time in response to demand. This individualised approach, based on allotting space and time to each vehicle, would in theory be far more efficient than the standardised signal phase and timing approach used for traffic lights today.

Automated, flexible traffic flows

The authors of the study recognise that an auction-based approach may not necessarily be socially acceptable, as it means allowing those who are in a hurry to pay in order to enjoy a faster journey. However, when it comes to finding a rational overall solution to traffic congestion their model has much to recommend it.  Leaving aside the question of whether and how to charge for this type of road system, one could still envisage automated regulation of city traffic flows according to the spaces available. Such optimisation could also help the driver – or driverless car passenger – to estimate his/her journey and arrival time. Given a direct two-way relationship between the traffic signal system and the in-vehicle software, the authors envisage the traffic signal system directing the car to take an alternative route on the basis of information relayed from the various sets of traffic lights if traffic is particularly congested in a certain area. The Madrid researchers argue that this approach could seriously improve overall urban traffic efficiency and significantly reduce journey times.

By Quentin Capelle