Bhibhya Sharma of the University of South Pacific, headquartered in Fiji, who led the research, said that the series of equations would instruct robotic cars when and how to merge lanes, which could lead to fewer accidents and ease traffic congestion.
Flocking is a biologically inspired technique and a strategy commonly used in robotics. "One of the advantages of flocking is that robots can work together and achieve what would take individuals far longer," he said.
He further explained that each car will be controlled by a centralized brain guided by a series of algorithms to control the car. The brains will talk to each other and instruct the cars to merge lanes and move in formation together.
The team demonstrated the technique using computer simulation and is currently testing it in two wheel robots.
Fellow researcher Utesh Chand said that the cars have targets they move towards and maintain and when the cars find themselves in a merging situation, one of them will be given the position of leader, the rest will follow.
The team has written equations for attraction towards the target, enabling cars to stay inside their lanes and avoid crashing into each other. The challenge now is to allow the robot to automatically plan how to avoid crashing with the side of the road and cars travelling close by.
Sharma and Chand presented these findings at the first Rim Mathematical Association (PRIMA) conference in Sydney, said a University of South Pacific release.
0 comments:
Post a Comment