Google has taught its self-driving cars to detect cyclists on the road, and has revealed that the use of sensors and machine learning has enabled its cars to interpret cyclists’ hand signals.
The company has stated that its self-driving technology is ready for public roads, as it can detect cyclists in a parallel-parked car with an open door. In this situation, it is programmed to slow down or nudge over to give the rider enough space to move towards the centre of the lane and avoid the door.
“We also aim to give cyclists ample buffer room when we pass, and our cars won’t squeeze by when cyclists take the centre of the lane, even if there’s technically enough space,” according to the Google Self-Driving Car Project Monthly Report.
Furthermore, the technology’s sensors have evolved to detect a rider’s hand signals in advance of a turn. The software has also learned to remember previous gestures to make better inferences about future riders.
Google ran a number of tests involving cyclists. One test involved its self-driving car cautiously approaching a cyclist that veered into its lane, and then stopping to avoid another cyclist that suddenly turned a corner and rode directly towards the car, against the flow of traffic. According to Google, the car was able to adapt to this unusual situation and prevent a collision.
The company’s self-driving technology uses a combination of cameras, Lidar and radar to collect information about objects that surround the vehicle.