Stuck in a jam? Not for much longer…

Have you ever been stuck in a traffic jam that seems to have appeared from nowhere, only to discover when you finally start to move again that there was no obvious cause?

Waiting in traffic queues could soon be a thing of the past, with new measurement technology giving us more information than ever on traffic flow along our busiest roads. Data collected from the active traffic management system now in place on the M42 in the West Midlands, will be analysed by Martine Barons and Dr. Stefan Grosskinsky in the Centre for Complexity Sciences at the University of Warwick who hope to use mathematics to get to the root of these mystery hold-ups.

Ms. Barons, a research student at the University of Warwick said “The thing that really excites me about mathematics is its extraordinary power to improve the lives of everyone, with applications ranging from traffic jams to disease control. Traffic jams are not merely inconvenient delays, but they waste finite resources, are bad for the environment and for health, and cost British industry up to £20 billion per annum, according to the CBI in 2003. “

These kinds of jams are called stop and go waves. Traffic comes to a sudden halt and drivers are stuck in slow moving traffic. After a while, they are able to speed up again, but have not noticed any reason for their delay. Stop and go waves travel in the opposite direction to traffic flow, and can be miles wide. The width is entirely dependent on the rate of inflow of traffic; the rate of traffic leaving is the same for every wave. It is believed that the trigger for these queues is lane-changing at congested merges. It is hoped that investigation of the M42 data will be able to confirm this and improve our understanding of the causes.

The M42 Active Traffic Management system was introduced between junctions 3a and 7 in 2006. This 17km section of motorway has a particularly complex pattern of traffic flow; in addition to cross-country traffic, it also provides access to the nearby city of Birmingham, the NEC and Birmingham International Airport. The scheme allows optimum traffic speeds to be calculated and implemented via overhead matrix boards. It also allows use of the hard shoulder as an additional lane during busy periods.

Inductance loops, are metal coils that are embedded in the surface of the road to detect the presence of cars as they pass overhead. A single pair of loops is used at pedestrian crossings or light-controlled junctions. For motorway traffic management, loops are spaced at regular intervals. The different loops are used together to gain more complex information on the speed and length of vehicles and on the flow rate of traffic in each lane at any given time. This allows the pattern of traffic movement to be recorded, so that when tailbacks do occur, the cause can be investigated. Dr Eddie Wilson from the Bristol Centre for Complexity Sciences, highlighted the importance of this traffic monitoring system. “It is the recent introduction of the availability of data that is transforming traffic modelling. The M42 active traffic management system data is particularly good because the detectors are closer than usual, at about 100m apart, with some just 30m apart. This enables us to build up a much more detailed picture of the traffic patterns and how they arise.”

Scientists at the University of Warwick are using mathematical models to simulate traffic flow based on information collected from these inductance loops. The movement of individual vehicles will be considered, taking into account variation in driver behaviour and the probability that a car will be able to move forward in different situations. This is dependent on the immediate environment of the car – whether there is space in front to move into and if so how much. The more free road ahead, the higher the probability that the car will continue to move forward. This type of analysis may hold the key to accurate predictions of when and where these queues will build up.

If this is successful, the group will be able to produce simulations that reflect the reality on the road and the system could be used to regulate traffic flow based on the prediction of areas where congestion is likely to occur. Dr. Stefan Grossinsky who is leading this work emphasised its importance. “The key to dissipating and preventing traffic jams is to have a clear understanding of how and when such patterns arise.”

The M42 Active Traffic Management scheme has been very successful in reducing queues by controlling traffic speed and lane usage in response to current traffic measurements. If successful, it is possible that this project may form the basis of another very effective traffic management scheme in the future, helping to wave a final goodbye to rush hour traffic jams.