Robot development system and method

The robot development system addresses the complexity of cerebellar control systems by training in virtual environments and integrating real-world data, improving the efficiency and adaptability of robot development.

TW202632527APending Publication Date: 2026-08-01DELTA ELECTRONICS INC(CN)
0 Cites 0 Cited by

Patent Information

Application Number
TW114121981
Authority / Receiving Office
TW · TW
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-01-17
Filing Date
2025-06-12
Publication Date
2026-08-01

AI Technical Summary

Technical Problem

The development of cerebellar control systems for robots is complex due to the need for high integration and adaptability in motion coordination, error correction, and motor learning, involving mathematical modeling, learning architecture design, instant reaction, and balancing accuracy and error factors.

Method used

A robot development system comprising a cerebellum controller, virtual component module, virtual construction module, virtual testing module, compilation module, and physical construction module, which includes training a virtual robot in a virtual environment, compiling the cerebellum controller into firmware, and constructing a physical robot, with optional physical testing and retraining based on real-world data.

Benefits of technology

This approach accelerates the development of robots by reducing the number of operations required and shortening the development time through automated compilation and integration of pre-built components, enhancing the adaptability and stability of cerebellar control systems.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

A robot development system is provided. The system includes a cerebellum controller, a virtual component module, a virtual building module, a virtual testing module, a compile module, and physical building module. The virtual component module stores a plurality of virtual components and at least one physical model corresponding to each virtual component. The virtual building module uses the cerebellum controller and the physical models of the chosen virtual components to build a virtual robot. The virtual testing module trains the virtual robot in a virtual environment to optimize the cerebellum controller’s control toward the virtual robot’s virtual components. The compile module compiles the trained virtual robot’s cerebellum controller into a firmware. The physical building module builds a physical robot according to the virtual components used by the trained virtual robot, and deploys the firmware to the physical robot.
Need to check novelty before this filing date? Find Prior Art