Skip to content

Reactive Driving Support - CrowdBot Release

Latest
Compare
Choose a tag to compare
@DrDiegoPaez DrDiegoPaez released this 17 Feb 15:53
· 2 commits to crowdbot since this release

This first release includes the main Reactive Driving Support (RDS), a method for robots to avoid imminent collisions with moving obstacles, which has been developed as part of research at LASA, EPFL.
It is included documentation for its simple implementation through ROS wrapper that includes methods for both non-holonomic robots, as well as, holonomic robots. The main input is obstacles' location and velocity which can be pre-processed dynamic obstacles tracking (e.g., yolo people detector) or raw proximity data (e.g., lidar point-clouds projected in 2D).
The main package was developed in C++, and has been tested with the robot Qolo with an average frequency of 150Hz for 2 LIDAR inputs in a single core i7 2.2 Ghz.