ITRI Takes on Manufacturing Challenges with RobotSmith

The Industrial Technology Research Institute (ITRI), Taiwan’s premier high-tech applied research institution, announced at CES 2023, RobotSmith, an artificial intelligence (AI) and robotic system for metal workpiece grinding and polishing.

At CES, ITRI’s highlighted technologies in AI, robotics and ICT. Among them is AI Aquarium, the world’s first smart aquarium that helps identify marine life in real time; Cubot ONE: Indoor/Outdoor AMR, an autonomous mobile robot (AMR) integrating AI, IoT and 5G technologies, that can operate indoors and outdoors, take elevators and travel on rugged ground; and RobotSmith.

AI algorithms are adopted to overcome bottlenecks of manufacturing processes.

Automates Metal Workpiece Grinding, Polishing

The newly developed AI and robotic system was specifically designed for metal workpiece grinding and polishing. ITRI developed RobotSmith to address the manufacturing industry’s challenges and trends of digital transformation, supply chain restructuring and labor shortages caused by the COVID-19 pandemic. RobotSmith’s quality grinding and polishing technology has been applied to manufacturing plumbing fixtures, high-end knives, and medical devices such as artificial joints.

RobotSmith performs high-quality surface finishing.

Specifically, RobotSmith digitalizes the grinding and polishing techniques of experienced workers. Also, it removes the limitations of human labor and manual precision. It is easily deployed in traditional industries to implement digital transformation while preserving craftsmanship skills required for production lines. Moreover, RobotSmith employs AI software to mimic human skills and overcome process bottlenecks.

RobotSmith incorporates sensing, robotics, industry best practices and ITRI’s AI software EzSim to offer a total solution for surface finishing. Specifically, EzSim software performs simulation and trajectory planning. It optimizes path parameters to achieve high precision and efficiency to perform surface treatment as delicately as a human expert. Moreover, sensors and trained AI models provide real-time defect detection and ensure process quality.

The system performs trajectory planning and path parameter optimization online and precisely controls the robotic arm on site.