Advancements in Humanoid Robot Performance Evaluation

Strategic Insights -

Research conducted by Jian Li and colleagues at the Beijing University of Posts and Telecommunications has resulted in a significant breakthrough in the development of a multimodal humanoid robot performance evaluation system. This system, based on a single-channel tactile-sliding triboelectric nanogenerator (SCTS-TENG), enables precise evaluation of humanoid robots' comprehensive performance, ensuring stable operation in various applications. The research has been peer-reviewed and published in Advanced Functional Materials, 2025.

Key Takeaways:

  • The proposed system uses a SCTS-TENG to enable 1D tactile position recognition, pressure perception mimicking human skin during sliding, and direction identification, achieving 97.22% accuracy as a standalone modality.
  • The system is constructed using the SCTS-TENG and a camera, with testing specifications defined for rotational contact of the dexterous hand and contact-sliding-separation of the robotic arm.
  • An in-depth analysis is conducted on output signal characteristics under different influencing variables, with the SCTS-TENG signals converted into 2D images using the gramian angular difference field (GADF) method.
  • The fusion of camera images and SCTS-TENG signals achieves 99.03% recognition accuracy using a multimodal deep learning (DL) system based on the Visual Geometry Group (VGG)19 model.
  • This system provides a low-cost, self-powered, and high-precision solution for sensing and evaluation of humanoid robots, offering significant advantages over existing methods.
  • The research has been conducted in collaboration with Gefan Yin, Runxiao Liu, Xiangyan Zhang, Yadong Mo, Linkun Zhou, Sikai Wang, Yuan Liu, Shimin Wei, Xuexiu Liang, Xiangjie Xu, and Zisheng Guo.

Statistics:

  • The proposed system achieves 97.22% accuracy as a standalone modality.
  • 75% accuracy is achieved by camera images as a standalone modality.
  • The fusion of camera images and SCTS-TENG signals achieves 99.03% recognition accuracy.
  • The system provides a low-cost, self-powered, and high-precision solution for sensing and evaluation of humanoid robots.

Sources:

  • A Single-channel Tactile-slip Triboelectric Nanogenerator for the Intelligent Performance Evaluation of Humanoid Robots. Advanced Functional Materials, 2025.
  • Wiley-v C H Verlag Gmbh, Postfach 101161, 69451 Weinheim, Germany (Wiley-Blackwell - www.wiley.com/; Advanced Functional Materials - onlinelibrary.wiley.com/journal/10.1002/(ISSN)1616-3028)
  • NewsRx. Findings on Androids Reported by Investigators at Beijing University of Posts and Telecommunications (A Single-channel Tactile-slip Triboelectric Nanogenerator for the Intelligent Performance Evaluation of Humanoid Robots). Journal of Engineering. October 20, 2025; p 950.