Intersections are dangerous places. Vehicles congregate in a centralised area going in opposite directions, creating opportunities for human error and accidents. But what will intersections look like in the age of autonomous vehicles? That’s the question a team from the University of Texas is seeking to answer.
Their concept, Autonomous Intersection Management (AIM) seeks to create a communication system between autonomous vehicles and the intersection itself. It works like this –
A vehicle approaches an intersection. The system that is manning the intersection communicates with the vehicle and identifies a suitable route. The section is duplicated many times over as different vehicles approach and leave the intersection. In essence, the system is acting as an air traffic controller except the entire behaviour is autonomous and managed by computers from start to finish. The result, when seen on a computer simulation, is a consistent flow of vehicles in an efficient fashion. In saying that, it’s also very scary to see cars fearlessly lurching into a crowded group of cars and popping out the other side.
This approach to the issue of autonomous intersections has peaked a significant amount of interest, with other universities seeking to create similar systems. In the past, autonomous vehicles were usually regarded as the solution to all problems, with sensors and communication systems speaking to other vehicles, but not to the road infrastructure. This new system means that an additional layer of communication exists which creates a new level of safety. With a centralised communication hub (which could be expanded to include far more than just an intersection) it’s not only the flow of traffic that will be enhanced. Safety warnings, mechanical failures and standard issues such as flat tyres, can be responded to at a macro level. The centralised communications hub can speak to other vehicles in the area, assist them and taking relevant action. For example, if vehicle A experiences a mechanical fault which forces it to slow down, it won’t only be the first vehicle behind it that will be made aware of the issue through external sensors – routing systems can use various models to ensure the flow of traffic is maintained. It’s an interesting idea and one that is likely to create a whole new layer of complexity for the growing ecosystem of autonomous vehicle creators.