Collision-Free Target Grasping in Complex and Cluttered Environment for Efficient Task and Motion Planning

This paper presents a novel algorithm that enhances the efficiency of robotic manipulation by enabling the robot to reach its target with fewer obstacles being removed, thus leading to faster performance times. The challenges faced by robotic manipulation in complex and cluttered environments (e.g., supermarkets, factories, etc.), where changes are constantly occurring, are well-known. To tackle these challenges, the robot should generate a collision-free path to reach the target while requiring minimal obstacle relocation to improve task efficiency. To achieve this, the proposed task planning algorithm incorporates the Rapidly-exploring Random Tree (RRT) algorithm and a modified Vector Field Histogram Plus (VFH+). The algorithm was tested in simulated service domain environments with objects placed on shelves followed by real-robot experiments. In both experiments, the results of the proposed algorithm demonstrated a lower number of obstacle removals (up to 37%) and consequently faster performance times compared to other existing methods (improved by up to 36%). Thus, the results showed that the proposed algorithm reduces the number of obstacle relocations and improves performance time compared to existing methods. As a result, the proposed method was evaluated as an improvement in task efficiency over previous approaches and has significant implications for the use of robotic manipulation in service domains with cluttered environments.