Researchers have shown that autonomous cars that compete with each other can speed up traffic 35%. The team from the University of Cambridge programmed small fleet miniature robotic cars to drive around a multi-lane track and observed the car's behavior when a car was stopped mid-circuit.
RELATED: THE 6 LEVELS OF AUTONOMOUS DRIVING AND THE FUTURE OF AUTONOMOUS CARS IN CHINA
Initially, before the cars were introduced to working together with the stopped vehicle cause others to slow down and wait for a gap in traffic before overtaking - a situation we often see in our own lives. A queue quickly formed behind the halted vehicle, and the flow of the circuit was slowed down.
Communicating cars made life smoother
But once the cars were communicating with each other the traffic flow increased. The son as a car stopped in the inner lane it sent a signal to all the other cars on the circuit. Vehicles on the outer circuit close to the stopped car, slowed slightly to allow for the inner circle cars to merge seamlessly minimizing the reduction in flow.
In the next step, a human controlled car was added to the circuit and driven in an ‘aggressive’ manner. The robotic cars could communicate to best to avoid the erratic driver.
The full results of the study will be presented at the International Conference on Robotics and Automation (ICRA) in Montréal.
Future streets will be safer
The information collected during the research will be integral to autonomous vehicle makers to better create ways for self-driving cars to communicate with human-driven vehicles.
“Autonomous cars could fix a lot of different problems associated with driving in cities, but there needs to be a way for them to work together,” said co-author Michael He, an undergraduate student at St John’s College, who designed the algorithms for the experiment.
“If different automotive manufacturers are all developing their own autonomous cars with their own software, those cars all need to communicate with each other effectively,” said co-author Nicholas Hyldmar, an undergraduate student at Downing College, who designed much of the hardware for the experiment.
The research was a clever inexpensive way to test different models. Communication between both autonomous cars and their peers and well as autonomous cars and their human-driven counterparts is a key area of interest for the industry.
But doing a real-world test is often difficult and expensive. The Cambridge students used inexpensive scale models of commercially-available vehicles with realistic steering systems, that were attempted with motion capture sensors and a Raspberry Pi, so that the cars could communicate via WiFi.
A lane changing algorithm developed for autonomous cars was then adapted for the fleet. The algorithm decides when a car should change lanes, based on whether it is safe to do so and whether changing lanes would help the car move through traffic more quickly.
This was then adapted to allow cars to be packed more closely when changing lanes and adds a safety constraint to prevent crashes when speeds are low.
A second algorithm allowed the cars to detect a projected car in front of it and make space. The cars were then tested in two driving modes, ‘egocentric’ and ‘cooperative.’
The researchers observed the reaction of the models on the stopped car in both modes and discovered cooperative driving improved traffic flow by 35% over egocentric driving, “Our design allows for a wide range of practical, low-cost experiments to be carried out on autonomous cars,” said Prorok.
“For autonomous cars to be safely used on real roads, we need to know how they will interact with each other to improve safety and traffic flow.”