Autonomous planning based on spatial concepts to tidy up home environments with service robots

Tidy-up tasks by service robots in home environments are challenging robotics applications because they involve various interactions with the environment. In particular, required not only to grasp, move, and release objects but also plan order positions for placing objects. this paper, we propose a novel planning method that can efficiently estimate of be tidied up learning parameters probabilistic generative model. The model allows robot learn distributions co-occurrence probability places tidy using multimodal sensor information collected Additionally, develop an autonomous robotic system perform tidy-up operation. We evaluate effectiveness proposed experimental simulation reproduces conditions Tidy Up Here task World Robot Summit 2018 international competition. results show enables successively several achieves best score among considered baseline methods.