Self-Guiding Laser Cutting Uses Sound To Reduce Metal Waste
Laser cutting offers speed, flexibility and high precision. It can also create complex metal parts in small batches. However, manufacturers must often run many tests before cutting a new alloy or material thickness.
Researchers at Empa and SUPSI want to simplify this process. Together with machine manufacturer Bystronic, they have developed a system that monitors each cut. It uses microphones, cameras and machine learning. The laser can then change its settings without human input.
Listening To Laser Cutting
High-powered lasers can cut through steel plates up to 25 mm thick. The laser melts the metal almost instantly. A stream of gas then pushes the molten material out of the cut.
Yet thick sheets can make the process less stable. Each alloy also responds differently to heat. As a result, manufacturers must adjust the laser for every new material. Even a small change can require a fresh set of tests.
The new system gives the machine both ‘eyes’ and ‘ears’. SUPSI focuses on camera monitoring, while Empa studies the sound of the cutting process. Although lasers use light, cutting metal creates a great deal of noise.
Empa recorded this noise with nine microphones inside the machine. The team tested steel sheets with thicknesses of 6, 10, 15 and 25 mm. Researchers then checked each cut by hand and used the results to train machine-learning models. They also ran smaller tests with several alloys.
According to Empa, sound can measure cut quality almost as well as cameras. Microphones also cost much less than advanced camera systems.
Cutting With Less Waste
The system checks the roughness of each cut edge. It also looks for burrs. These deposits form when molten metal sticks to the back of a sheet. Thick sheets often produce more burrs because they need greater laser power.
Fewer burrs mean less grinding and other finishing work. This saves time and energy. It can also reduce rejected parts and material waste.
These benefits could support more efficient metal production. The technology may interest automotive, product and architectural designers who use precision-cut metal components.
Upgrading Existing Machines
The partners aim to complete the two-year project in autumn 2026. During the final phase, they will connect the sound and camera models to one control system. This system will check each cut and change the laser settings at once.
Manufacturers could also add the technology to existing machines. They would only need a computer, the control software, microphones and cameras. The system could therefore improve metal processing without requiring a completely new laser cutter.
Source: Empa
Photos: Roland Richter / EMPA