An autonomous robot is useful and successful if it is able to accomplish the desired actions in the right sequence and properly react to environmental changes. The mission system (or Artificial Intelligent (AI) software module) of such vehicles is responsible for decision making. Modifications to this system are time-consuming and error prone since code modifications, recompilations and executable uploading are industry-used standards. The SONIA Autonomous Underwater Vehicle (AUV) team from Ecole de technologie superieure (ETS) in Montreal, Canada has developed a proof-of-concept demonstrating the advantages of on-the-fly reconfigurable mission system. The mission system is composed of an on-board mission framework built around the simplest AI system, a state-machine based system. The AI system is fully reconfigurable using a telemetry interface featuring mission saving, mission loading, state reconfiguration inside a mission, state transition management based on conditions, parameters assignation and model checking to ensure mission integrity.
展开▼