Many SMEs are still reluctant to use artificial intelligence (AI). An initial starting point is to analyze your own value chain: Where and how can AI provide support as a powerful computing tool? In this interview, Prof. Dr. Alexander Löser explains how SMEs can benefit from AI, which process steps need to be taken and why the introduction of AI requires a willingness to fail. He is the founder and spokesperson of the Data Science Research Center at Beuth University of Applied Sciences Berlin and a member of the "Technological Pioneers and Data Science" working group of the Learning Systems Platform.

Many SMEs still have some catching up to do when it comes to digitalization. To what extent is AI already an issue for them?

Prof. Dr. Löser: That depends on the value chain to which an SME wants to contribute or which it may already control in part or in full. It always becomes exciting when an SME has already 'occupied' part of the value chain and wants to expand to the 'left' or 'right'. An example from medicine: hospitals currently control the flow of patients as soon as they arrive at the emergency room. But why do they appear there? This is where the providers of apps or devices that people use to record their symptoms come into play. The app advises them to visit the emergency room or book a telemedical consultation if they have these or those symptoms, now or later. In other countries - with less regulation and healthcare services that are too expensive for many people - the app can also advise them to 'just' buy a medicine. The provider of this app thus manages patients along the entire value chain - from the assessment of symptoms to diagnosis, diagnosis, treatment, rehabilitation and lifestyle. This can lead to 'traditional' medical service providers losing revenue or to certain, for example lucrative, cases being diverted - depending on the margin planned by the provider of the healthcare app.

In Europe, we should research AI technologies for such applications - ideally with our strong German partners in the healthcare system - in order to ensure sufficient sovereignty in the future. The Data Science Research Center at Beuth University of Applied Sciences has proposed a corresponding use case for the European cloud alternative GAIA-X: deep patient representation for differential diagnosis. We are happy to present our work in clinical deep learning to interested partners.

Turning unused data into added value - know how!

Manufacturing SMEs have previously unused data at their disposal. How can this be used to create value-adding data products and services?

Prof. Dr. Löser: There are numerous tried-and-tested methods and "construction kits" that we use ourselves in consulting with our partners. One of the first steps is to analyze the value chain and identify worst-case scenarios, pain points and key figures, e.g. sales, growth, customer satisfaction, which need to be optimized for the company. The next step also has nothing to do with AI: Defining the target markets. Are we talking about one hundred to one thousand customers for a data product or platform - or hundreds of thousands? Is it a transaction-based business model or do we need to build a community for an innovation platform and corresponding business models?

Only after it is clear where the exact and preferably scalable benefit for the top 5 data product ideas lies, can we move on to the technical implementation: What are the core business objects that we should represent in an AI system? Then we should start to translate the benefits in value creation into maximizable benefit functions. In the example mentioned at the beginning, for example, patients - and their symptoms - could be mapped to diseases and possible urgencies. This often results in a classification whose input variables are the business objects. Other methods include time series analyses, clustering or even multi-task learning. The strength of AI lies in its combinatorial power - i.e. the ability to quickly calculate the probability of the best matches in a smart way. Another trick is to strengthen the often incomplete representation. And finally, feedback loops need to be taken into account - after all, we want our model to continuously improve.

Research projects with industry

What challenges do SMEs face when introducing AI - and how can they tackle them?

Prof. Dr. Löser: The challenges are similar to many other projects: Data products are risky. We often only know whether a customer request is feasible once the training data is available. That's why the introduction of AI requires money, talent and time - around twelve months for the first data product if there is no experience yet. It also requires a willingness to fail and move on, as well as support and trust from top management. For many German companies, this is more of a marathon than a sprint. Many do not dare to do this or do not even see the starting points. Others have this on their radar, but rely on their well-running sales and production processes and therefore see no need for major seven or even eight-figure investments.

In the BMWi-funded Servicemeister project, the Data Science Research Center at Beuth University in Berlin is working with a portfolio of typical German companies such as Krohne, Wirth, Atlas Corpco, Trumpf and KEB, all of which have a very good understanding of AI mechanisms and value chains.Probably the biggest challenge lies in producing data products for small and medium-sized B2B markets. In terms of organization, this requires a smart approach to data engineering and as little dependency as possible on existing IoT platforms, which could drive up prices. Technologically interesting approaches are those in which the development process of an AI model is largely carried out by AI-based machines - without the need for expensive human data engineers.

In addition to basic research, the transfer of AI to companies is a crucial issue. It is precisely this duality that we need even more of in Germany. I would therefore very much like to see universities as well as universities of applied sciences, which are particularly strong in research, benefit from the funding for AI centers, which are currently doing outstanding work in basic research and the transfer of AI.