Old World Computing GmbH in Bochum is a data specialist. The team offers professional consulting and services relating to data science: training, specific data-driven projects and customised software solutions. However, it's not just technical expertise and an understanding of IT that are required for the digital transformation in companies to gain momentum.
What challenges do you face when working with your customers?
Sebastian Land: The use of AI is a fundamentally new approach because it is completely beyond our horizon of experience. Until now, we've only known how to make someone else think from humans, but not from machines. AI is a new tool that you really have to understand before you start working with it. The employees who manage AI-supported projects also need to understand it, because it has completely different requirements than a normal project.
Conceptually, we are facing something new that people are also afraid of. The media fuel this feeling. We can only counter this by building up knowledge. We educate people so that they understand what AI is and can do.
What contribution does your work make to enabling companies to operate more sustainably?
Sebastian Land: We are focussing on the area of industrial production. Essentially, the aim is to solve problems in which the computer learns independently. It learns to solve a highly complex problem - for example a control system in industry - better than humans can. At the moment, specialists are looking at 500 monitors, each with 10,000 numbers, turning cogs and - if we're honest - making decisions based on gut instinct. Nobody can check the settings.
We can teach the computer to deduce from historical data how people have acted and what the results have been. On this basis, the computer can make predictions as to which settings would optimise certain goals at a given time: Minimise energy use, minimise raw material use, maximise quality - these would be such classic optimisation goals. Humans can use the predictions to try out different parameter settings even before the product has been manufactured. The computer is simply better than us in this respect because it can process data faster and recognise patterns better.
AI is a tool
Which sectors are predestined to use AI?
Sebastian Land: In industry, but also in agriculture, there is already a large database - at least for larger systems - as the control system there is electronic. However, the data must also be stored. It is therefore more a question of mentality that is independent of the industry. With one exception: in the IT advertising industry, i.e. Google & Co, artificial intelligence is of course the be-all and end-all.
Your company is based in Bochum - can you find enough skilled labour?
Sebastian Land: Our team consists of computer scientists, mathematicians, engineers and economists and we are also feeling the shortage of skilled labour. But the Ruhr region in particular, with its universities, offers opportunities to find good people. Dortmund is the centre of excellence for AI in Germany. It leads the programme of six national competence centres for AI research, which are funded by the German government. As a student, I didn't realise this significance at first. It wasn't until I was a guest at a conference and a murmur went through the community when the professor with whom we had our lectures in Dortmund entered, that I realised it.
Sebastian Land
Founder and Lead Data Scientist
Sebastian Land (40) was born in Bochum, studied computer science at TU Dortmund University, worked there at the Chair of Artificial Intelligence and supported the spin-off of the start-up RapidMiner, where he later played a key role in the development of the platform, sales, support and customer consulting in various positions for over ten years. In 2015, he became self-employed and founded Old World Computing GmbH.
How can new technologies such as AI be used successfully in practice?
Sebastian Land: I have to think strategically about AI, but it's mostly tactical thinking. Some department has a problem and someone wants to use AI and then someone does it. But that's wrong. I have to learn to do this strategically as an overall organisation, then it will also be efficient. Small departmental approaches lead to a dead end because the necessary resources are also lacking. But above all, I also need to know: What am I getting myself into? Training an AI is like research or science. You don't know whether there is a solution or when it will materialise. It may not work. There is a probability, but no guarantee. It's a quest, a bet on the future. That's a mentality that we don't have. AI requires a completely different approach to projects.
What do you recommend to your customers?
Sebastian Land: It makes sense to set up your own science data team in order to build up and retain knowledge in-house. A small team that can grow and offer workshops and training. It is important to explain to employees that their work will not disappear because processes are supported by AI. There have to be people to replace the engine on the production line. AI can't do that. It can send electrical signals via data lines, but it cannot solve physical problems. There won't be a Terminator running around the world - that would be far too expensive.
Get in touch with us.
Do you have questions about knowledge transfer in the Ruhr region? Or are you interested in our networks? Give us a call or contact us directly using the contact form.