R&D Focus
How Robots Help Preserve and Improve Heritage Craft in Manufacturing Process By Darryl Tsai
Metal grinding and polishing factories such as plumbing fixture manufacturers require sophisticated craftsmanship to produce high-quality output. However, many manufacturers have long suffered from a shortage of experienced professionals, and require robotic assistance to help increase productivity.
To address this issue, ITRI has developed RobotSmith-EzSim, a software that incorporates Cyber-Physical System (CPS) for robotic applications to automate the output and hasten the process of human craftsmanship for metal parts manufacturing. CPS is based on 3D models of robots, peripheral equipment and workpieces to perform functions such as trajectory planning. The CPS system can enhance programming efficiency and flexibility to shorten the time for robot teaching and initial setups. This allows RobotSmith-EzSim to easily perform grinding and polishing for customized products.
A simulation software called RobotSmith-EzSim incorporates CPS to enhance robotic applications.
Our team has collaborated with HCG, a major Taiwanese manufacturer of bathroom fixtures to deploy this technology in their factory. During the cooperation, we applied RobotSmith-EzSim to mimic and perform skills of craftsmen to produce faucets with complex appearances. RobotSmith-EzSim successfully demonstrated its capability to achieve various production targets and satisfy the needs of the industry. This software can be applied in other sectors as well, for instance, aerospace manufacturing and semiconductor equipment manufacturing. It is suitable for use in fabricating aircraft components such as engine blades and those with irregular shape and curved surface.
In addition, we adopted vision-based workpiece recognition containing 3D visual orientation and trajectory compensation with laser scanning for workpiece measurement and precise positioning in RobotSmith-EzSim. This resolves the known issue that the use of virtual models in CPS may not always accurately reflect the real environment and minimizes the simulation-to-reality gap. During the grinding or polishing process, acoustic emission (AE) sensors are used with deep learning mechanisms to analyze frequencies and vibration for product quality monitoring. Hence, virtual grinding is added to assess and monitor outcomes of different grinding parameters like the angle or the contact force between the belt and workpieces to improve quality control.
An image of RobotSmith-minicube.
A new robotic system RobotSmith-minicube was first introduced at the TAIROS 2020 in Taipei with these developed functionalities. Driven by RobotSmith-EzSim, it demonstrated attractive refined grinding and polishing capabilities including contour measurement and vision recognition. Furthermore, RobotSmith participated in ITRI Startup Day and was awarded as a finalist.
The RobotSmith Team was recognized as a finalist on ITRI Startup Day.
As a digital master, RobotSmith presents a cutting-edge solution to deliver advanced automated applications in grinding and polishing processes, while functioning to preserve heritage craft in the manufacturing sector.
About the Author
Darryl Tsai is an associate researcher at ITRI. He received both his master’s degree and bachelor’s degree in Mechanical Engineering at National Chiao Tung University (NCTU). His research interest centers on the development of Cyber-Physical Systems for industrial robot applications.