AI-assisted laser material processing using fast compact piezo-driven 3D beam shaping and new laser sources (AI-BEAM)
Dynamic beam modification in combination with multi sensor concepts and artificial intelligence control algorithms allow a paradigm shift in laser macro material processing
Sensor-supported laser material processing with three-dimensional dynamic beam shaping
In a holistic solution approach, laser source, three-dimensional beam shaping and process sensor technology are coupled for efficient and reliable process control for laser cutting, welding and hardening. The dynamic three-dimensional beam shaping enables a temporal and spatial change of the energy distribution in the range of several kilohertz, whereby this can be adjusted sensor-controlled within a few milliseconds.
The low latency requires an enormous amount of computing power, which is why multi-sensor monitoring of laser processes and AI-based data processing are the subject of current research in order to develop a new generation of inline process control. The long-term goal is to take into account further necessary part and production information in order to develop concepts for cloud-based data-driven services.
Project Results
As part of the project, an FPGA-based infrastructure for real-time data acquisition was developed, integrating acoustic sensors, high-speed cameras, and light-section sensors into a single measurement platform. In highly dynamic welding and cutting processes, multimodal sensor data with high temporal and spatial resolution was acquired inline, synchronized, and preprocessed for AI-supported analysis. The information obtained initially enables automated quality decisions for processes and components. In the future, it will be used, for example, to adaptively adjust three-dimensional beam shaping to the process within a few milliseconds, thereby enabling intelligent inline process control.