Obstacle Avoidance for Human-Centric Environments Using a Modified Artificial Potential Field - Probabilistic Roadmaps Based Planner

Serial manipulators are essential for automating tasks and working alongside humans in the fields of automotive, pharmaceutical, retail, and research. However, there is a significant challenge to ensuring both human safety and efficient operation. This paper tackles this challenge by introducing an improved approach that combines the Modified Artificial Potential Field (APF) method with a Probabilistic Road Maps (PRM) planner, utilizing an optimization-based pose estimation model approach to represent human obstacles using RGB and depth information. The proposed method is compared with the commonly used Rapidly Exploring Random Tree (RRT) and Probabilistic Road Map (PRM) for path planning in environments where humans are involved. The study focuses on a six-degree-of-freedom serial manipulator wall-mounted on a flat surface. By comparing the proposed approach with existing planning algorithms, the paper analyzes the developed Modified APF-PRM and its performance based on metrics such as path length, variation of joint angles, and time taken for path planning. The paper concludes with a discussion of potential improvements and refinements for algorithms based on the results of this study.