Carnegie Mellon University (CMU) scientists have developed a new algorithm to help autonomous vehicles (AVs) safely navigate narrow, crowded streets.
The research was led by CMU’s Argo AI Center for Autonomous Vehicle Research. The team includes Christoph Killing – now part of TU Munich’s Autonomous Aerial Systems Lab, research scientist John Dolan and Ph.D. student Adam Villaflor.
Killing and his colleagues first developed a method to model different levels of driver cooperativeness – how likely a driver was to pull over to let the other driver pass – and used those models to train an algorithm that could assist an autonomous vehicle to safely and efficiently navigate this situation.
Though the algorithm has only been used in simulation and not on a vehicle in the real world, the results are nevertheless promising.
What’s more, the team found that their algorithm performed better than current models.
Killing believes their research is the first to deal with this specific driving scenario. It requires drivers – human or not – to collaborate to make it past each other safely without knowing what the other is thinking.
Driving down a crowded, narrow street can be quite challenging even for seasoned drivers. Excepting an autonomous vehicle to do this has been next-to-impossible so far.
Drivers must balance aggression with cooperation. An overly aggressive driver, one that just goes without regard for other vehicles, could put itself and others at risk.
Similarly, an overly cooperative driver, one that always pulls over in the face of oncoming traffic, may never make it down the street.
Add parked cars that line both sides of the road to the mix, and you have a nightmare on your hands.
One has to basically duck into a gap in the parked cars or slow and pull over as far as possible for the other to squeeze by.
“It’s the unwritten rules of the road, that’s pretty much what we’re dealing with here,” says Killing.
“It’s a difficult bit. You have to learn to negotiate this scenario without knowing if the other vehicle is going to stop or go.”
Driving is full of complex scenarios like this one, adds Dolan. A solution in this regard could bode well for autonomous vehicles tasked with last-mile deliveries as it could enable them to navigate tight spaces and unknown driver intentions.
Image and content: Carnegie Mellon University
