Development and Optimization of a Cost-Effective Lightweight Humanoid Robotic Arm for Assistive Applications

Abstract

This paper presents the development of a cost-effective, lightweight humanoid robotic arm designed to assist the elderly and vulnerable populations. The study aims to provide specialized robotic arms by utilizing a motion planning method based on human arm biomechanics. The arm, created using 3D printing technology with 40% infill, achieves a weight reduction of over 60%. Low-torque servos and a human-like adaptable gripper further enhance cost efficiency and functionality. The arm features simplified joints and is driven by six modified R/C servomotors with analog feedback for precise angle measurement. System identification shows high accuracy, with joint fit percentages ranging from 87.5% to 97.07%. A PID controller, optimized via genetic algorithm (GA), particle swarm optimization (PSO), and honey badger algorithm (HBA), ensures rapid and accurate positioning. The Simscape library models the arm dynamic behavior, addressing forward and inverse kinematics, workspace, and path planning. These innovations promise to advance assistive technologies significantly.