Lyft announced that it has started to leverage data from its ridesharing network to advance and train its self-driving systems.
A select number of drivers’ cars are now equipped with camera sensors to capture difficult situations and help generate 3D maps. Lyft notes that the data it collects will help it to update its 3D maps of roads and buildings.
The sensors also give information about speed limits and pedestrian crossings. The company is able to generate this data by using AI to identify traffic objects to understand how risky driving situations are handled.
“We focus on data and objective metrics for safety, and pay special attention to vulnerable road users,” the company wrote in a blog post.
The rideshare company says that it has provided drivers in the program information about the camera it is using and the information that it is collecting. It notes that the camera doesn’t collect audio and that it doesn’t face the driver.
“Lyft uses a holistic and iterative systems safety approach to ensure the safety of our testing program, which includes simulation, closed-course testing, and on-road operations,” it noted.
Lyft outlines that it believes self-driving cars will be a critical component of a safer transportation system, and that it is committed to the safety of all road users including bikers and pedestrians.