Behaviour-Based Docking using the DAMN Arbiter
T.E.R. Bailey, E. Nebot · 2008
This paper discusses a method that enables reliable docking of a mobile robot with a rectangular container. The process involves an aligned a oach to the container while avoiding any obstacles in the region and avoiding collision with the container itself. A modular behaviour-based architecture called the Distributed Architecture for Mobile Navigation (DAMN) is used. Raw sensor data is processed to produce robust sensors that provide input data to the behaviour modules. Centralised arbiters then asynchronously process the behaviour outputs and detennine the set points for the mobile robot drive and steering actuators. Testing of the virtual :sensors and behaviour-based algorithms was perfqrmed on an indoor mobile robot, SydNav, with wheel-encoders and a scanning range laser as its sensors. Further testing of a container poseestimation sensor took place on a quayside cargohandling vehicle (a straddle-carrier); again using scanning lasers for its sensorial input.