테이블 내 빈 공간 확보를 위한 식기 쌓기 작업계획

Abstract

In situations where dishes are densely arranged on a table, robotic hands face challenges in effectively manipulating objects due to limited space and obstruction caused by surrounding items. This study addresses these challenges by proposing a task planning algorithm to create available table space by stacking dishes. The proposed algorithm builds a Knowledge Base (KB) using the Vector Field Histogram (VFH), which analyzes the spatial distribution and density of surrounding objects, and defines a heuristic function and a pruning method based on the KB. The Heuristic function is integrated with a backtracking search method to quickly generate a local solution that satisfies the target density. Experimental results demonstrated that the algorithm could reduce table dish density (rho) from 0.5 to 0.35 within 1 second, making it practical for real-world applications.