EAISI Director Carlo van de Weijer is a busy man. He’s just on his way from his car to his office on TU/e campus when he takes our call. On the occasion of its first lustrum, the Eindhoven-based AI institute is organizing a symposium. “I was just called by national daily De Telegraaf,” Van de Weijer tells us while he’s looking for a room to talk freely. “When it comes to AI, the media have started to find their way to us as well.”

More importantly, the same applies to students and companies. “Five years ago, if you wanted to study Artificial Intelligence (AI), you wouldn’t end up at our university of technology,” Van de Weijer reminisces. “The business community was demanding for us to participate in research projects. And it still is, but now we can meet those demands. Students can study AI here, and we are launching one research project after another in collaboration with the business community.”

Multidisciplinary research

The investment budget of 100 million euros has been spent. “We mainly used it to hire PhD candidates and other academic staff,” says Van de Weijer. “We funded lots of multidisciplinary research and set up many laboratories.”

Those academic faculty aren’t employed by EAISI, but by various departments. AI can be applied to any technological domain. “Researchers were often hired with our funding, but entered the employment of a department. Now the first investment program is over, they remain with that department. They pay for themselves by securing their own research projects.”

Cross-fertilization

EAISI brought about cross-fertilization and mutual collaboration, Van de Weijer says. For example, at the Department of Electrical Engineering and that of Mechanical Engineering people turned out to be working on autonomous driving without being aware of one another’s existence. “Good intentions abounded, but if you put those people together, you’ll get much better ideas.”

As an academic discipline, AI is in its infancy. And yet, something of a specialization is starting to emerge, Van de Weijer indicates. “AI research at TU/e is mainly focused on health technology, mobility and high-tech industry.” Each academic environment chooses its own research domains. For example, in Amsterdam people are mostly working on things like language and writing recognition, as well as AI applications in the financial sector, says Van de Weijer.

‘Cyber only’

What makes AI research at TU/e unique is that applications manifest themselves in the real world. “For a long time, AI was a ‘cyber only’ discipline,” Van de Weijer explains. “Thanks to AI, Spotify knows exactly what the next song should be. But if that goes wrong, it doesn’t matter. However, if AI doesn’t navigate your car onto the offramp but into the divider, you do have a problem.”

This makes AI in the TU/e domains extra exciting, Van de Weijer thinks. By now, there are about 900 people at the university that work in AI, across the different departments. “The investment program is now continuing with funding from ‘strategic means’, but an increasingly large portion is funded from external monies.”

AI students

Alongside research, more and more AI education is taking place as well. EAISI’s goal was to take in 300 bachelor’s students each year, supplemented by 300 master’s students and an additional 300 people from the business community that receive further training in the field of AI. “We’ve achieved that goal, but the bar is – or rather: should be – raised. AI programs will serve as an important pillar in the growth or our university,” Van de Weijer says in closing.