New mathematical technique could reduce traffic jams with autonomous, connected cars
06 November 2013
Researchers have developed a mathematical technique that allows the calculation as to whether there is a chance a traffic jam will happen from a uniform flow of traffic. A paper on the work by Dr. Róbert Szalai at the University of Bristol (UK) and Dr. Gábor Orosz at the University of Michigan is published in the journal Physical Review E.
City traffic is made up of cars that are similar but not identical. Computer-driven cars are also similar, but they can react faster and, through wireless communication, observe traffic not just in front of them.
Current understanding of traffic is somewhat limited to identical cars, so to advance technology and design safe robotic drivers the researchers needed to understand models including non-identical participants.
The study has found that by using the design tool and controlling the reaction times of robotic cars this can reduce the possibility of traffic jams and accidents.
City traffic is made up of cars that are similar but not identical. Computer-driven cars are also similar, but they can react faster and, through wireless communication, observe traffic not just in front of them.
Current understanding of traffic is somewhat limited to identical cars, so to advance technology and design safe robotic drivers the researchers needed to understand models including non-identical participants.
The study has found that by using the design tool and controlling the reaction times of robotic cars this can reduce the possibility of traffic jams and accidents.
The analytical nature of our research means it could be used to understand all sort of systems, such as to design driverless cars that could work well in a mixed human-robotic environment or control gene regulation in living organisms, for example suppress the expression of some harmful gene.
—Dr. Róbert Szalai
The researchers hope to develop in the future more realistic car-following models and connectivity structures. It is hoped that by using similar mathematical equations the research can be used to understand of all sorts of systems from robotic/human cars in traffic, gene regulatory networks and neural networks that could explain epileptic seizures.
Resources
Róbert Szalai and Gábor Orosz (2013) “Decomposing the dynamics of heterogeneous delayed networks with applications to connected vehicle systems,” Physical Review E doi: 10.1103/PhysRevE.88.040902
Nice idea, sort of AI "super cruise" control.
Posted by: SJC | 09 November 2013 at 04:23 PM