Robot applications are being increasingly used in real life to help humans performing dangerous, heavy, and/or monotonous tasks. They usually rely on planners that given a robot or a team of robots compute plans that specify how the robot(s) can fulfill their missions. Current robot applications ask for planners that make automated planning possible even when only partial knowledge about the environment in which the robots are deployed is available. To tackle such challenges we developed MAPmAKER, which provides a decentralized planning solution and is able to work in partially known environments. Decentralization is realized by decomposing the robotic team into subteams based on their missions, and then by running a classical planning algorithm. Partial knowledge is handled by calling several times a classical planning algorithm. Demo video available at: https://youtu.be/TJzC_u2yfzQ.
MAPmAKER: Performing multi-robot LTL planning under uncertainty
Pelliccione P.
2019-01-01
Abstract
Robot applications are being increasingly used in real life to help humans performing dangerous, heavy, and/or monotonous tasks. They usually rely on planners that given a robot or a team of robots compute plans that specify how the robot(s) can fulfill their missions. Current robot applications ask for planners that make automated planning possible even when only partial knowledge about the environment in which the robots are deployed is available. To tackle such challenges we developed MAPmAKER, which provides a decentralized planning solution and is able to work in partially known environments. Decentralization is realized by decomposing the robotic team into subteams based on their missions, and then by running a classical planning algorithm. Partial knowledge is handled by calling several times a classical planning algorithm. Demo video available at: https://youtu.be/TJzC_u2yfzQ.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
