DEXTEROUS MANIPULATION

Dexterous manipulation is a critical capability for robots operating in complex, unstructured or dynamic environments. It enables robots to handle a wide variety of objects and tasks with precision, adaptability, and autonomy.

A*STAR focuses on developing intelligent manipulation systems that combine tactile sensing, learning-based control and motion planning to perform fine, contact-rich interactions with the environment. Integrated with real-time perception, these systems can operate effectively in dynamic settings, extending their reach and flexibility. The systems also incorporate intuitive human-in-the-loop control to support human safety and efficient semi-autonomous operation in challenging scenarios. 

Dexterous manipulation
This image was generated by Microsoft Copilot.

KEY RESEARCHERS

  • Dr. Wu Yan, A*STAR Institute for Infocomm Research (A*STAR I2R)
  • Dr. Liang Wenyu, A*STAR Institute for Infocomm Research (A*STAR I2R)
  • Dr. Acar Cihan, A*STAR Institute for Infocomm Research (A*STAR I2R)
  • Dr. Zhao Xinyuan, A*STAR Institute for Infocomm Research (A*STAR I2R)

KEY PROJECTS

  • Dexterous Robot Manipulation for Delicate Precision Engineering Processes
    • This collaborative project with a precision engineering MNC aims to develop tactile-driven perception and control algorithms for robotic systems to accurately perceive their environment, perform adaptive stable grasping and dexterous manipulation. The robotics hardware system will undergo iterative enhancements based on performance validation in real-world precision engineering task scenarios.

Contact Dr. Wu Yan to learn more.

  • Advancing Safe Human-Robot Interaction: An Integrated Approach for Enhanced Perception, Communication, and Control (SafeHRI)
    • The SafeHRI project aims to enable service robots to be socially compatible in real-world public spaces through the development of intrinsically safe control systems. 
    • A*STAR developed a robust control framework that ensures robots operating in dynamic, human-populated environment to remain safe and stable. This includes reactive strategies for handling sudden environmental changes, maintaining operational integrity without compromising task goals. The architecture is designed to be modular and integrable with commercial robotics perception and planning systems.

Contact Dr. Wu Yan to learn more.