Learning Feasibility for Task and Motion Planning in Tabletop Environments
Learning Feasibility for Task and Motion Planning in Tabletop Environments
复制标题
DOI：
10.1109/lra.2019.2894861
复制
发表时间：
2019-04-01
影响因子：
5.2
通讯作者：
Kavraki, Lydia E.
中科院分区：
文献类型：
--
作者：
Wells, Andrew M.;Dantain, Neil T.;Kavraki, Lydia E.
任务和运动规划(TMP)结合了离散搜索和连续运动规划。早期的工作表明，为了有效地找到任务运动计划，离散搜索可以利用关于连续几何的信息。然而，将连续要素融入离散计划器带来了挑战。通过将几何知识以稳健的方式引入到基于约束的任务规划器中，提高了TMP算法在具有固定机器人的桌面场景中的可扩展性。其关键思想是学习可行运动的分类器，并使用该分类器作为启发式来排序搜索任务-运动计划。学习的启发式方法将搜索引向可行的动作，从而减少了动作规划尝试的总次数。我们方法的一个关键特性是允许在不同的场景中进行强有力的规划。我们在最小样本场景上训练分类器，然后使用原则近似以最小化错误影响的方式将分类器应用到复杂场景中。通过将学习和计划相结合，我们的启发式方法在不同的桌面场景中产生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可以确保我们的计算代价很小。
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 table-top 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 toward 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 table-top scenarios. Even when classification errors are present, properly biasing our heuristic ensures that we will have little computational penalty.