Lyft is an America based ridesharing company. The services are currently active in more than 600 cities in the United States and 12 cities in Canada. Lyft, through its mobile app, offers services for cars, scooters, and bicycle-sharing systems. It also runs a food delivery service.
Currently, Lyft is working on an ambitious project of building a self-driving car. The data isn’t collected from all but a selected number of vehicles in its Express Drive Program. The program provides rental cars and SUVs to drivers as an alternative to a long-term lease.
How is the data collected?
It has collected data from the Lyft drivers to boost the development of the project. The level 5 programs aim at studying 3D maps, understanding human driving patterns, and from the results obtained, improve simulations. All these tools are necessary to make a modern vehicle technology.
The Express Drive vehicles are installed with forward-facing camera sensors for collecting the data. The project is also gathering data from its autonomous vehicles as well as the safety cars. The drivers are given detailed information regarding the camera and the data being collected. There are no microphones to collect audio in the cameras, and they aren’t aimed at the driver in any way.
Blue Vision Labs technology is being used to convert driver data into city-scale 3D geometric maps. Keeping the 3D geometric maps up-to-date and scaling their scope is really challenging. Lyft’s officials in a blog said that they had mapped thousands of miles due to their coverage.
The cameras also collect data on the various situations a driver faces daily. These situations are studied and then used to make more realistic and sophisticated simulations.
The trajectory of the road is tracked by visual localization technology. It is used to maintain the best location in the lane, which always may not be the center lane.