Text Steffen Maltzan ––– Photography
Artificial Intelligence (AI) enables machines to imitate human capabilities such as logical thinking, learning, planning and creativity. As a result, technical systems can be aware of their surroundings, respond to them, and solve problems independently. Have you ever heard of software that uses generative AI (GenAI) to create new content independently to achieve goals? How about machines that pay attention and think actively? GenAI is an attractive proposition for every area of automation. It is therefore no real surprise that the forecasts regarding economic potential are huge. In June 2023, McKinsey predicted that the productivity boost provided by GenAI could contribute 2.6 to 4.4 trillion dollars of added value a year to the global economy. However, what elements have already been sufficiently tested that they are ready for industry adoption?
AI ALONE IS NOT ENOUGH
“Alongside information management, the most obvious area of application for AI in production facilities is visual quality control,” says Dieter Meuser, CEO of German Edge Cloud. Cameras have been used for this for a long time. AI is now taking this to a new level. While the traditional approach involves getting data scientists to train visual models and involves a lot of effort, AI support makes it possible for departments and workers to do this themselves. “Such applications have long since proved themselves in proofs of concept,” says Meuser. “However, they need to meet even more requirements if they are truly to take their place in everyday work on the shopfloor.” Meuser can see two main focal points: “It has to be simpler to connect them to existing systems. And it has to be straightforward for users to correlate the data from quality control with the data from the production processes.”
German Edge Cloud is now making that possible with ONCITE Analytics VisionAI, a new service from Digital Production Systems in cooperation with IBM. The modularity of the digital production system’s microservice-based software architecture cuts the work and cost involved in the implementation, while at the same time ensuring digital continuity. By linking up with the operational data from ERP and machine data from the shopfloor, ONCITE DPS creates the perfect platform for transparency, insights and traceability in production processes. When the results of quality assurance checks are also added into the equation, transparency is expanded further still, opening up the possibility of more productivity optimisation.
THE FUTURE IS GENERATIVE AI
“The many medium-sized manufacturing companies out there are in particular need of solutions that can be implemented quickly and are cost-effective and scalable with just a few users as soon as they are launched,” says Meuser. “By combining our DPS with IBM software for visual inspections, we are able to offer them this way in, with a solution from a leading software provider.” For Meuser, it is already clear what the next trend will be. “Generative AI will become more dependable as a useful tool for industry. When it comes to visual inspection, the AI would then be capable of generating its own simulated error images with which to train the models, thus further increasing the speed and quality of detection quite significantly.”