Combined task and motion planning as classical AI planning
Planning in robotics is often split into task and motion planning. The task planner decides what needs to be done, while the motion planner fills up geometric details. However, such a decomposition is not effective in general as the symbolic and geometrical components are not independent. This dissert...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis doctoral |
| Estado: | Versión publicada |
| Fecha de publicación: | 2018 |
| País: | España |
| Institución: | CBUC, CESCA |
| Repositorio: | TDR. Tesis Doctorales en Red |
| OAI Identifier: | oai:www.tdx.cat:10803/563078 |
| Acceso en línea: | http://hdl.handle.net/10803/563078 |
| Access Level: | acceso abierto |
| Palabra clave: | Combined task and motion planning Task planning Motion planning Heuristic planning State constraints Classical planning Robotics Modeling task and motion problems Combinació de planificació de tasques i moviments Planificació de tasques Planificació de moviments Planificació heurística Restriccions d'estats Planificació clàssica Robòtica Modelar problemes de tasques i moviments 62 |
| Sumario: | Planning in robotics is often split into task and motion planning. The task planner decides what needs to be done, while the motion planner fills up geometric details. However, such a decomposition is not effective in general as the symbolic and geometrical components are not independent. This dissertations shows that it is possible to compile combined task and motion planning problems (CTMP) into classical planning problems; i.e., planning problems over finite and discrete state spaces with a known initial state, deterministic actions, and goal states to be reached. Motion planners and collision checkers are used for the compilation, but not at planning time. What makes our approach effective is 1) a fully compilation of CTMP problems into classical planning problems, 2) expressive classical planning languages for representing compiled problems, using functions and state constraints, 3) general planning algorithms capable of finding plans for CTMP problems using domain-independent heuristics. |
|---|