- Education
- 02/12/2025
“AI will fundamentally transform education at TU/e”
Banning it no longer makes sense, and pretending nothing is happening doesn’t either: AI is already having a disruptive impact on education at TU/e, and that impact is only going to grow, says professor Tom van Woensel. He has been tasked with embedding AI properly into all aspects of education within the university.
How can you be sure an assignment was not made with the help of AI? You can’t. How can you prevent students from using AI in their studies? You can’t. And as for the development of artificial intelligence itself, which is moving at breakneck speed: that certainly cannot be stopped. That is why TU/e felt it was necessary to provide clarity on the use of AI in education at the university, both for students and for lecturers, and to develop a clear vision for it.
To do this, the university called in the help of Tom van Woensel. He is a full professor and program director at Industrial Engineering and Innovation Sciences (IE&IS) and has recently become AI in Education portfolio lead. His task is to steer a transformation that should ensure that by 2030 AI is seamlessly integrated into education at TU/e.
Oil tanker
To achieve this, entire educational programs and assessment methods, among other things, will need to be reconsidered. That is not something that happens overnight, Van Woensel knows. “I see it as an oil tanker: to change course even a little, you need hundreds of kilometers.” In the short term, this process starts with providing students and instructors with tools, for example in the area of assessment. For some projects or forms of examination, the use of AI may be permissible, but when it comes to properly assessing knowledge, it may also be desirable that a student does not have access to it, Van Woensel explains. “For some exams, you may have to go back to pen and paper, or to an oral format.”
According to Van Woensel, there is no point in testing students on simple, easily searchable facts anymore. That information is, after all, instantly available through AI. “Instead, we need to focus on fundamental knowledge and complex issues that cannot simply be solved by AI.” Challenge-based learning and working in a multidisciplinary way could help with this, he believes. In the long run, AI could actually support that latter aspect. One of the university’s goals is to use artificial intelligence to enable personalized learning.
Full freedom
Van Woensel points to EPFL, a university in Switzerland, as an example of where this is already happening. “They have developed a tool they call ‘Graph Search.’ They have processed all the information from all the courses at the university into a kind of searchable environment.” The tool pulls relevant information about a specific topic from all those different courses. “Suppose you want to know something about integral calculus. You need limits, derivatives, and calculus for that. You see those connections in a kind of network.”
For TU/e, this is still a distant dot on the horizon, but it could enable the ultimate form of personalized learning, in which a student has full freedom in shaping their study program. “As long as, at the end of the day, you can demonstrate to instructors or a committee that you have achieved the learning objectives at a sufficient level.” Van Woensel acknowledges that this is an extreme example, but it does show what AI could mean for the university.
There are already occasional critical voices within the university about the wide range of choices students currently have. So what happens if that freedom becomes even greater in the future? “I hope that in the future we will recognize that there are different learning models that we as a university can facilitate.”
Skilling up
Back to the near future: first, Van Woensel believes that TU/e needs to develop more awareness around the use of AI in education. Awareness of the possibilities, but also of the problems it can create. The working group is exploring ways to add an AI literacy component to the learning paths for personal and professional development for students. Lecturers could also use some upskilling in this area, he notes. “They are the ones who ultimately have to reform their programs.”
When it comes to the use of AI, Van Woensel believes that students are much further along than most lecturers. Students were not yet involved in the initial initiative for rolling out AI within education at TU/e, but he does want to involve them in every step going forward. “For every element we are now elaborating, we have included roles for students, who can be brought in via different councils and committees.”
University Council
Last week, the University Council approved a “Framework for the Development of AI in Engineering Education,” which Van Woensel drafted together with the working group established for this purpose. The working group includes, among others, the dean of the Bachelor College, Ines Lopez Arteaga, and several program directors.
When approving the document, the education committee of the University Council did raise a few concerns. According to its members, the framework is not clear enough about which additional AI skills engineers should master. They also said that the document was paying too little attention to the ethical and environmental implications of the large-scale use of AI, and that they felt the working group had involved students too little so far. Rector Magnificus Silvia Lenaerts acknowledged these three points and promised to take them into account more in the future.
Discussion