日本語 
Mitsubishi Electric is pushing deeper into smart manufacturing with a new edge digital twin technology built for CNC machine tools. This is not theory work sitting in a lab. It is aimed directly at fixing a very real shop-floor problem.
In collaboration with RWTH Aachen University, the company has developed a system that detects and corrects machining errors as they happen. The focus is on small deformations caused during cutting. These are easy to miss but costly when they stack up across production.
The setup runs on edge computing with high-speed processing. It continuously collects data like axis positions, currents, and cutting forces at a high sampling rate. Instead of overloading the system, it filters out only what is needed to estimate errors. That data feeds into a compact physical model, built using a minimal set of equations, and the output is sent straight back into the CNC control system in real time.
Also Read: OKI Launches Complete EMS for AI Server Equipment
The impact is straightforward. Testing shows error reduction of up to 50 percent. That means fewer defective parts, more consistent surface quality, and less rework. And that is where the real value sits. Better productivity without increasing resource use, and less waste generated during manufacturing.
Step back and this fits into a larger shift. Digital twins have advanced from their original function as simulation tools to their new role as operational control systems. The system not only detects future issues but also resolves them during active machine operation.
English