Robots assume that a synchronized car without a driver can increase traffic flow

Science

New research shows that synchronized cars can reduce driverless traffic and increase traffic flow by up to 35 percent.

To test the potential of automotive-less-automotive technology, Cambridge University researchers built a fleet of small robot cars. The cars are placed on a mini test drive. The researchers measured the impact of vehicle damage on the rest of the fleet.

If vehicles function independently of each other, stopping the vehicle causes a significant reaction. However, when the car has worked in a coordinated manner and communicates with each other while driving on the track, stopping in the lane will only cause a short delay.

When a synchronized car collapses in the inner lane, the robot vehicle in the outer lane slows down to facilitate the merging of cars in the inner lane.

Synchronization – this is a vehicle that communicates with each other – increasing traffic by 35%. The researchers shared the results of their tests on Monday at the International Robotics and Automation Conference, which took place in Montreal this week.

“Autonomous cars can solve various driving problems in the city, but there must be a way to work together,” said Michael Heed, a Cambridge University student.

He helped develop algorithms that allow communicating cars to synchronize their actions in response to damage.

Researchers have used motion sensors and Raspberry Pi to let their miniature robot cars feel their environment and communicate via Wi-Fi. In order for cars to coordinate their actions, he and his research partners have adapted algorithms that allow real non-driver cars to make safe belt changes.

The effort is to test the technology of vehicles without drivers or based on computer simulations or real vehicles. However, using a real car is expensive and requires more space. Researchers from Cambridge suggest that their method can make testing automotive technology easier and cheaper without a driver. Each of their cars costs around $ 75.

“Our design enables a variety of practical and cost-effective experiments for autonomous vehicles,” said Amanda Nabi from Cambridge computer scientists. “In order for cars to be used safely on real roads, we need to know how they interact with each other to improve traffic safety and flow.”

Leave a Reply

Your email address will not be published. Required fields are marked *