Wells Andrew M, Dantam Neil T, Shrivastava Anshumali, Kavraki Lydia E
School of Engineering, Department of Computer Science, Rice University, Houston TX andrew DOT wells AT rice DOT edu, lastname AT rice DOT edu.
School of Engineering, Department of Computer Science, Colorado School of Mines, Golden CO dantam AT mines DOT edu.
IEEE Robot Autom Lett. 2019 Apr;4(2):1255-1262. doi: 10.1109/LRA.2019.2894861. Epub 2019 Jan 23.
Task and motion planning (TMP) combines discrete search and continuous motion planning. Earlier work has shown that to efficiently find a task-motion plan, the discrete search can leverage information about the continuous geometry. However, incorporating continuous elements into discrete planners presents challenges. We improve the scalability of TMP algorithms in tabletop scenarios with a fixed robot by introducing geometric knowledge into a constraint-based task planner in a robust way. The key idea is to learn a classifier for feasible motions and to use this classifier as a heuristic to order the search for a task-motion plan. The learned heuristic guides the search towards feasible motions and thus reduces the total number of motion planning attempts. A critical property of our approach is allowing robust planning in diverse scenes. We train the classifier on minimal exemplar scenes and then use principled approximations to apply the classifier to complex scenarios in a way that minimizes the effect of errors. By combining learning with planning, our heuristic yields order-of-magnitude run time improvements in diverse tabletop scenarios. Even when classification errors are present, properly biasing our heuristic ensures we will have little computational penalty.
任务与运动规划（TMP）将离散搜索与连续运动规划相结合。早期的工作表明，为了有效地找到任务-运动计划，离散搜索可以利用有关连续几何形状的信息。然而，将连续元素纳入离散规划器存在挑战。我们通过以稳健的方式将几何知识引入基于约束的任务规划器，提高了固定机器人在桌面场景中TMP算法的可扩展性。关键思想是学习可行运动的分类器，并将该分类器用作启发式方法来对任务-运动计划的搜索进行排序。所学习的启发式方法引导搜索朝着可行运动的方向进行，从而减少运动规划尝试的总数。我们方法的一个关键特性是允许在各种场景中进行稳健规划。我们在最小示例场景上训练分类器，然后使用有原则的近似方法将分类器应用于复杂场景，以尽量减少误差的影响。通过将学习与规划相结合，我们的启发式方法在各种桌面场景中实现了数量级的运行时间改进。即使存在分类错误，适当地偏向我们的启发式方法也能确保我们几乎不会有计算代价。
