The German Research Center for Artificial Intelligence is now also represented at the University of Trier
Since this week, the University of Trier has officially become a branch of the German Research Center for Artificial Intelligence (DFKI). The focus of future joint research will be on the intersection of artificial intelligence (AI) and business informatics. The two professors of business informatics at the university, Ralph Bergmann and Ingo Timm, spoke with the “Luxembourg Word” about their research plans.
Ralph Bergmann, Ingo Timm, what does it mean for the University of Trier to be a branch of the DFKI?
Ralph Bergmann: First of all, it means that we, as scientists, have the opportunity to build dfki groups. This enables us to carry out projects that are carried out by the DFKI and benefit from the infrastructure, at the premises of the University of Trier. The advantage for us is that we can use the contacts of the DFKI and thus also have the opportunity to transfer our previous research at the university into practical applications much more strongly. The focus will be on intersections between AI and business informatics.
Where are these points of contact?
Ingo Timm: Over the past 15 years, we have found that there is an increasing need to design complex flexible systems in the economy that must also be able to deal with new situations – i.e. not all solutions that have already been prepared in the code. The need for technical systems capable of processing information, i.e. to gain some understanding of it, is very high in business informatics. The systems should be able to assess specific situations and then make decisions on the basis of the given framework.
Can you give a concrete example of this? Ingo Timm: For example, the operational planning of nurses. What happens if a nurse falls ill or a new need arises for the carer? An alternative must be searched for in the planning algorithm accordingly.
Joint research is supposed to be about experience-based systems, among other things – what is it?
Ralph Bergmann: We develop AI applications that are able to use the knowledge that can be gained from data to solve problems. We humans do the same. We are able to solve new problems based on our experience and learn and get better with increasing experience. Experience can be used, for example, in medicine Ralph Bergmann (l.) and Ingo Timm. in the form of medical cases, i.e. patients who show certain symptoms, who have been found to be the cause of a particular disease and have been treated appropriately, the success or failure of which has been established. This is an example of a medical Experience that can be used in diagnosis and therapy and thus evaluated new situations due to similarity. It is always a question of using specific experiences to solve new problems.
Why do you need AI when people also learn from experiences?
Ralph Bergmann: The physician, for example, has his own experience. Ideally, a computer system allows us to bring together the experiences of different physicians and use them more systematically. We can use this experience from many sources, which goes well beyond what individuals can do. A second aspect is cognitive social simulation. What is this all about?
Ingo Timm: If we look at the changes and aspirations within The Framework of Industry 4.0– for example, that the person should be in the foreground and determine the rhythm of production – then we have concrete questions.
How can a person be involved in the planning process as a customer or service provider?
Our planning systems usually see people as a kind of resource that is available at certain times in the shift schedule or not. But when we look at the efficiency of processes, it is also much more important to understand the incentives that affect people, the extent to which they are motivated to participate, what types of skills they have to make decisions, and when they make decisions. what they might look like. The same question applies to AI. Now we need to consider how this changes the process, how people interact with AI systems, for example, and how to separate the function between humans and systems.
What about the fear of many people that AI could put jobs at risk for real people?
Ingo Timm: From an expert’s point of view, this fear does not arise at all, because we build systems that experts in their everyday practice. Of course, there will be automations in which a system can take over, but we assume that it will create a space in which man can fill the actual task he has even more meaningfully. Especially in the medical field, we currently have a lack of resources, there are too few people available for the tasks involved. And that, of course, can relieve the ai. On the other hand, we also have new processes. AI is often very specialized, but never a unit that can completely replace humans. The image of the drug detection dog fits well. We have a highly specialised system that may even work better than a human in an area, but only in a very small field. The moment we almost stop looking for the drugs the dog is targeting, it becomes difficult. And you still need experts who can do the job.
Are other research areas involved in the work in addition to business informatics?
Ingo Timm: Of course, we incorporate the expertise of other subjects, such as from the legal field, psychology, health care or social statistics. We have many cooperation partners to work together on a project-by-project basis.
Do you also have cooperations in Luxembourg?
Ralph Bergmann: At the moment we do not have any active cooperations. We were in discussion with colleagues from the List (Luxembourg Institute of Science and Technology, ed.), but we do not yet have a concrete project. However, we are also seeking cooperation with the regional economy here in the Trier area, but also, of course, in Luxembourg. For example, smaller projects and feasibility studies could be carried out, which, if funded by companies, for example, can also quickly produce new results.
What would you say, is the aim of your research?
Ralph Bergmann: Essentially, it is project-like activities in which one tries to solve an application problem together with an application partner. It is not the case that software is created, which is then sold off the shelf. The questions we deal with are far too individual. The aim is, and this is also the main reason why we have sought to cooperate with the DFKI, to implement projects in which our research results are incorporated into practice and thus have a truly practical effect. The aim is to implement projects in which the results of our research are incorporated into practice. Ralph Bergmann