There’s little doubt that driverless cars are on their way. And as organisations around the world plot for their arrival, researchers are working hard to anticipate and solve some of the challenges self-driving vehicles will bring.
One such problem is that of how driverless cars will be able to mix with manually-driven cars on the roads – especially when joining existing lanes of traffic.
Now, researchers in Switzerland believe they may have the answer.
Merging driverless vehicles in traffic
The team from the country’s École polytechnique fédérale de Lausanne (EPFL) has developed an algorithm that they say will allow automated vehicles to operate in traffic alongside manually-driven ones.
Other approaches have focused on the concept of vehicle platooning – for example, having a convoy of self-driving trucks follow a lead one (typically manually-driven). However, because each convoy operates as a single block, managing the convoy gets harder as it grows.
The EPFL team, working as part of the European AutoNet2030 project, has taken a different approach. Instead of a single leader guiding the other vehicles, the team’s system harnesses inter-vehicle communication to allow each of the vehicles to communicate with others nearby – adjusting their speed and position independently.
By doing so, the convoy can travel on one or more lanes on a highway, and can automatically rearrange itself when a vehicle leaves the group or a new one joins. There is also no upper limit to the convoy’s size, says the EPFL team.
EPFL’s Distributed Intelligent Systems and Algorithms Laboratory (DISAL) developed the base algorithm for the software that controls the convoy.
“We have been working on this type of distributed control algorithm for around ten years,” said Alcherio Martinoli, the head of DISAL. “Simply put, the idea is to find a way for agents that are not particularly clever – robots or cars – to work together and achieve complex group behavior.”
Real-life testing on autonomous vehicles
The DISAL research project progressed in steps, starting with the team managing robots on simulators. They then moved on to real life miniature robots, before going back to the simulators to test out their algorithm on cars. Finally, they were able to test it using real vehicles on the road.
The team conducted a demonstration on a test track in Sweden involving an automated truck and car, plus a networked car that was driven manually. This car was equipped with GPS and laser sensors, as well as a human-machine interface to allow the driver to follow instructions when joining the convoy.
“It may not seem so impressive with only three cars, but for the first time we were able to validate what we had achieved in the simulation. And the number of vehicles in the convoy has no impact on the complexity of the control mechanism,” said Martinoli.
Guillaume Jornod, the EPFL scientist who ran the trials, noted that the test was a proof of concept.
“We are hoping that, with a rise in demand, carmakers will come up with ever cheaper solutions for converting legacy vehicles, that they will coordinate their efforts with the community working on the Internet of things, and that we will be able to deploy and improve this multi-lane convoy system for heterogeneous vehicles,” said Jornod.