Car manufacturer Ford is to release a ‘comprehensive self-driving vehicle data’ to researchers in an attempt to promote further research and development of autonomous driving technology, according to its Autonomous Vehicle Manager, Tony Lockwood.
Announced via a blogpost by Mr Lockwood, the package includes data from multiple test vehicles operating in Detroit in the US and collected over a one-year period. The testing that generated this data was separate from work Ford is doing with Argo AI to develop a production-ready autonomous driving system.
Available data includes: Camera and lidar sensor data, GPS and trajectory information, multi-vehicle data, a 3D-point cloud, and ground reflectivity maps. A plug-in adds a visualisation element for the data, which is offered in ROS format, Mr Lockwood said. The fact that the data was collected over the course of a year, means it includes seasonal variations accounting for weather, traffic and pedestrian density, and construction, he said.
Mr Lockwood said Ford is the first to offer data from multiple vehicles. Having more than one vehicle operating in the same general area provides different snapshots of data, since one vehicle’s sensors might capture something another’fs do not, Lockwood said. Working with these different pieces of data could prove useful to researchers, he indicated.
Researchers will also be able to use Ford’s data as a baseline against their own work, or to construct simulations for testing new autonomous driving software, according to Mr Lockwood .
The data will be available online through Ford’s collaboration with the Amazon open data program. The first data set has already been uploaded, with more to come, Mr Lockwood confirmed.