Researchers build self-driving cars that can navigate without maps
Researchers at the Massachusetts Institute of Technology’s (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) have built a system that allows self-driving cars to navigate on roads they have not been on before without 3D maps. Current self-driving cars rely on maps to determine where they are and how they can react to obstacles, but such maps are not available for less travelled roads. To tackle this issue, the CSAIL team has developed the MapLite system which combines GPS data with sensors that observe the road conditions. The system relies on sensors for all aspects of road navigation and uses GPS data only to obtain a rough estimate of the car’s location. After a final destination and a ‘local navigation goal’ (which has to be within the view of the car) are established, the sensors generate a path to get to the destination, by estimating the location of the road’s edges. As explained by CSAIL, ‘MapLite can do this without physical road markings by making basic assumptions about how the road will be relatively more flat than the surrounding areas’.