17 Jun AI triggers innovation boost in photonics
Advancing digitalization and sensor-based process monitoring provide raw data that can be used to generate information with added value. This accelerates research and development projects, saves iteration loops in the design of optics and beam paths and, in combination with virtualization supported by artificial intelligence (AI), reduces the number of real experiments.
Photonics-driven advances in sensor technology, imaging and data processing are paving the way for real-time process monitoring as the basis for AI-based on-the-fly adaptation of process parameters, for example. This brings the vision of self-learning machines and ultimately closed control loops in laser processing closer.
Manufacturers and integrators are using AI for automated component optimization and for more efficient, more productive and less failure-prone process chains in laser cutting, bending and welding. Neural networks filter previously inaccessible information from large amounts of data. In ongoing laser processes and in quality assurance, AI supports the detection, segmentation and evaluation of hidden defects and anomalies in camera or computer tomography images. In clinics and laboratories, it ensures fast, reliable findings.
AI in the life sciences
Microscopy providers use the technology along the entire process chain: AI helps to place and recognize samples, supports focus adjustment and image acquisition, corrects noise when microscoping living cells and automates 3D reconstruction, image segmentation and object classification. Scaling leads to more reliable findings in research and diagnostics, as AI-based analysis is based on much larger data volumes.
AI for laser technology
SMEs such as Precitec and Scansonic say they are well advanced when it comes to AI-supported laser welding, laser soldering and the accompanying measurement technology. “AI methods are changing everyday life in the photonics industry,” says Dr. Markus Kogel-Hollacher, Managing Director of the Laser Technology Working Group and responsible for research and development projects at the Precitec Group. According to Kogel-Hollacher, the potential of AI is particularly great in laser processing. This is because there are many uncertainties where photons penetrate workpieces and develop their effect. All key process input variables must be meticulously adhered to in order to rule out quality fluctuations. Sensor technology and AI help and expand the options: “It’s no longer just about good or bad, but with AI we elicit precise statements about physical properties such as the strength or contact resistance of a weld seam,” explains Kogel-Hollacher. In this way, AI-supported inline quality monitoring creates real added value in the laser welding of batteries, fuel cells or consumer electronics.
Photonics for smart factories and energy-efficient AI data centers
Christoph Franz, Managing Director at 4D Photonics, has a similar view: “The smart factories of the future will need lean, flexibly integrable solutions that enable AI-supported multi-sensor process monitoring and evaluation.” AI filters relevant information for process improvements from the generated sensor data. To generate this data, 4D Photonics relies on multispectral imaging. A single sensor provides the necessary data in an uncomplicated, fast and dynamic way to examine laser processes with the required precision and provide relevant information on the cause of errors in the process, explains Franz.
The interaction between photonics and AI goes beyond this: providers of integrated photonics, including Coherent and Q.ant, are driving forward purely optical computing processes and data transmission. “Our photonic processors replace electrons with light, making it possible to reduce energy requirements by a factor of 30 while increasing computing density by a factor of 50,” reports Q.ant Managing Director Dr. Michael Förtsch. A real paradigm shift. His conclusion: “A more scalable, more cost-effective and more energy-efficient AI future is possible”. Photonics is the way to get there.
Photonics meets AI and robotics
Laser World of Photonics 2025 in Munich will address the megatrend of AI in photonics in the application panel ‘From theory to practice: Implementation of AI in laser-based manufacturing processes’ on Tuesday, June 24, from 11:00 to 13:00 in Hall B3. There will also be a special themed show by Laser World of Photonics and Automatica with exhibits and presentations entitled ‘Photonics meets Robotics: AI Success Stories’ in Hall A3.
Source and image: messe-muenchen.de