Nieuws

New chair on AI between Atlas Copco Airpower and KU Leuven

Mensen ontvangen certificaat leerstoel

October 8, 2024

Atlas Copco Airpower and KU Leuven further strengthen their long-term cooperation on artificial intelligence with a new 4-year chair. The chair is another step in Atlas Copco Airpower's innovation strategy.

"In doing so, we are investing in the technology of the future. The possibilities of AI are far from exhausted in our industry, and we are constantly looking for new ways to innovate."

Wouter Ceulemans, President Airtec divisie

Artificial intelligence is playing an increasing role in the daily lives of a lot of people, think for example of Chat GPT, editing or facial recognition on photos, suggestions around interesting series or purchases or translations from a foreign language. AI is also playing an increasing role within the industrial sector. Some examples are automation through robots and machine data tracking for predictive and prescriptive maintenance.

At Atlas Copco Airpower's request, a research team from the DTAI department of the KU Leuven computer science department will now further investigate how artificial intelligence can provide solutions for industrial machines and processes. This will be led by Professor Jesse Davis, an expert in machine learning and AI. The goal is to use advanced AI models to improve processes and make them more sustainable

"Applications of artificial intelligence in the field of text and image, such as Chat-GTP, have become widely known over the past year and continue to continuously improve themselves thanks to the large amount of data. In industrial processes, the situation is different. The potential of AI in industry is great, and at Atlas Copco Group we are already making full use of it. But it also has its challenges. Due to the complexity of industrial machines and processes and more limited availability of data, the classic AI learning techniques we have today are reaching their limits. Therefore, we want to research newer learning techniques that are more advanced and better able to deal with those challenges."

Philippe Geuens, Teamleader engineering

Frontrunner in AI applications in industrial context

At Atlas Copco Group, for example, the lack of failing compressors and thus data about them, and the gradual degradation process of a compressor is a challenge for classical learning techniques of AI algorithms. More recent AI techniques do have the potential to understand how machines degrade, when service is needed and how they interact with other machines. Optimizations of AI through the Chair will allow for better insights, leading to even better products and service.

On top of that, AI solutions can also support in the production process by helping workers produce the large range and different variants of products more efficiently. For example, learning systems based on artificial intelligence could support workers to quickly and efficiently master production and assembly techniques, allowing them to handle a wider variety and thus adding variety to the job.

"At Atlas Copco Group, we invest in the technology of the future, and innovation is key. Innovation allows us to continue to grow as a company and deliver even more value to our customers. With techniques such as artificial intelligence, we continue to provide them with the best experience and solutions for their needs. We are constantly looking for ways to innovate, and the possibilities of artificial intelligence are far from exhausted. We are frontrunner in our industry in this regard."

Wouter Ceulemans, President Airtec divisie

KU Leuven as reliable partner in innovation

Atlas Copco Airpower and KU Leuven look back on decades of successful cooperation. For example, they have already worked together as part of the Chair in Service Systems, with KU Leuven conducting research into efficient and sustainable maintenance of industrial machinery.

"The potential of AI for industry is extremely high. But there are still many challenges, notably the lack of data and labels and the failure to fully exploit existing knowledge and documentation. This Chair will allow us to develop new techniques that meet these needs. For example, we will strengthen our research that focuses on AI learning techniques that combine data-driven and knowledge-driven methods, allowing flexible supervision. We are also looking at techniques that combine learning and reasoning. The goal is to improve integration between machine and expert and leverage model verification to increase the reliability of AI models."

Professor Jesse Davis, expert in machine learning and AI - KU Leuven

Leerstoel AI KULeuven 2024-trofee

© Michael De Lausnay

Atlas Copco Airpower donates this Chair about artificial intelligence in ight of the 600th  birthday of KU Leuven. More information: https://steun.kuleuven.be/hoe-steunen/leerstoelen/irw/atlas-copco-leerstoel-voor-onderzoek-in-artificial-intelligence