Prosthetics and Rehabilitation RoboticsRobotic Locomotion and ControlSoft Robotics and Applications

M. T. Ali, Abir Ahmed, Khaliduzzaman Mredul, Jagannath Bhowmik, Md. Abdur Rahman

2026.5.31AIUB Journal of Science and Engineering

DOI: 10.53799/deaa3j54

Abstract

The optimization of the actuation systems in the humanoid robotic arm is the focus of this research, which considers different factors including motor performance, selection of materials, and innovative mechanisms. By systematically testing a number of Brushless DC (BLDC) motors with different KV ratings and reduction ratios, it was concluded that the 5048 100KV motor with 30:1 cycloidal drive was optimal, with a high KV rating, high efficiency, low weight and low price. According to the voltage analysis, the 24V system doubled the torque, reduced heat generation and increased responsiveness over the 12V system. Comparisons of material properties between 3D-printed PLA+ and aluminum revealed that PLA+ could be used for lightweight prototype making and aluminum could be used for highly loaded parts due to high strength and high precision. The limitations that are faced by the conventional designs (high inertia, mechanical complexity, energy inefficiency) are overcome by proposing to move forward from the conventional design to a Low Inertia Manipulator with High Stiffness and Strength (LIMS). This is a tension amplified mechanism where the tension is amplified using cable driven mechanisms with high torque to weight ratios, low backlash and high compliance as compared to conventional BLDC based mechanisms. The results provide practical guidance in designing a lightweight and energy-efficient robotic arm and indicate that a hybrid approach for the prototyping to production process is necessary and LIMS should be used in the design of next generation humanoid manipulators.

Citation format

ALI, M. T., et al. Optimizing humanoid arm actuation: Motor selection, material trade-offs, and transition to low-inertia tension-amplified mechanisms. AIUB Journal of Science and Engineering, 2026, 24(2): 178.