At the AstaZero test track, a mock-up of a city junction has been used to simulate a real-world traffic environment with both autonomous and manually-driven vehicles negotiating with each other and adjusting their speeds in a cross intersection.
The team – made up of innovators and researchers from Ericsson, Chalmers University of Technology, the University of Naples “Federico II” and AstaZero – have used 5G cellular network technology and distributed cloud to exchange safety-critical data between both autonomous and manually-driven vehicles and the road infrastructure.
“Collisions are avoided by arranging the vehicles within a virtual platoon and enforcing inter-vehicle distances such that both side and rear-end collisions are avoided”, says Paolo Falcone, Associate Professor in the Mechatronics research group at Chalmers.
“Our task has been to develop algorithms for controlling the vehicles”, continues Paolo Falcone, who during the project has supervised a doctoral student and a master´s student from the University of Naples “Federico II”. “These algorithms have then been implemented on the vehicles by help of ReVeRe, Ericsson and AstaZero.”
Most modern vehicles already have the cellular network technology required to transmit information like position and speed data, but restrictions of traditional radio networks prevent this data from being used in safety-critical applications like avoiding collision.
By bringing the network much closer to the point of use and leveraging the low-latency power of edge computing, vehicles can communicate this data with each other rapidly and reliably, positioning themselves to avoid collision on the approach to a common intersection. This is opening the possibilities of a much smoother driverless transport network, as well as it is an excellent proof point for using the network in new ways. Not just to communicate, but to help us make better decisions and improve safety.
More about the research
The project was conducted from March to June 2018. The researchers had experience from a similar project
, but did everything from scratch since different control algorithms, communication technology and vehicle platforms were used.
For information, contact
, Associate Professor in the Mechatronics research group at the department of Electrical engineering at Chalmers