CKI Conference 2020

Like everything this year, the annual get-together of Siemens and RWTH Aachen University was a bit different than usual. For safety reasons, we held a virtual conference but at least some of our speakers and the moderator were able to join us in our little “studio”.

Key takeaways of the conference

• There is no “right” strategy for implementing AI – the best way is to establish ecosystems that work their way backwards from the objective to the research question.
• When it comes to autonomous driving, AI and mobility experts need to communicate and also include safety experts in their conversations.
• Intelligent algorithms are nothing without humans who control and interpret the processes and results.

Despite the circumstances, we had a great and interesting conference focusing on Artificial Intelligence in the contexts of Mobility and Production. As usual, we included both sides – researchers from RWTH Aachen and managers from Siemens. This way, we covered a lot of aspects and views of these topics.

First of all, our CKI Director Prof. Malte Brettel welcomed Dr. Peter Körte, the new CTO of Siemens AG as the new Management Sponsor of the CKI in Aachen. We are really happy to once again have an experienced and well-connected manager within Siemens on our side to initiate research projects and take the partnership to the next level. Dr. Körte mentioned the importance of collaborating since looking only inwards and taking care of yourself truly limits the potential for innovation.  

Trustworthy AI and the Importance of Ecosystems

We started off with Dr. Ulli Waltinger, Founder and Technology Head of Siemens AI Lab and Prof. Bastian Leibe, Head of Computer Vision at RWTH Aachen University, who introduced us to the complex world of Artificial Intelligence. Research should be driven by demand, meaning that we should start with the desired outcome and work our way backwards to determine what exactly needs to be developed. We need to work towards a trustworthy AI that is robust, explainable and secure. Especially the security and safety issues will make AI more acceptable with users. There is continuing progress in the capabilities of deep learning methods for Computer Vision.

Future Mobility in Smart Cities

We then introduced the topic of Mobility into the conference with Dr. Claus Bahlmann, Head of R&D Department Artificial Intelligence at Siemens AG and Prof. Lutz Eckstein, Director of the Institute for Automotive Engineering (ika) at RWTH Aachen University, who are both naturally using AI on a daily basis. While Claus Bahlmann at Siemens focuses on AI solutions for rail and train applications, Lutz Eckstein provided examples of current research in autonomous driving in street traffic.

Putting these four experts into one panel of course quickly led to a vivid exchange on the topics of autonomous driving and smarter cities. One key takeaway here is that in order to speed up the development in this area, AI experts from different disciplines need to work together more closely and also include safety experts in their planning. Only then can developers and providers be sure to find acceptance of autonomous trains and cars in our daily lives.

Digitalisation of Machine Tools

Next up, Prof. Christian Brecher, Chair for Machine Tools, RWTH and Michal Skubacz Head of Industry Software Motion Control, Siemens AG presented familiar topics to our viewers since the collaboration between those two institutions goes back quite a few years. Christian Brecher explained the way from raw data to knowledge and how important it is to have humans refine the data to have meaningful, usable AI in production processes. In order to realise the use cases that Christian Brecher showed during the presentation, Michal Skubacz demonstrated how SINUMERIK Edge can bring intelligent data directly into the machine for additional process insights and improved accuracy.

And again, our two very own AI experts Bastian Leibe and Ulli Waltinger then joined the two previous speakers and engaged in a lively discussion about the implications of algorithms in the production processes. One point we know for sure: With the growing impact of AI we need to make sure the decisions are fair and explainable. Nevertheless, we as humans have to understand the results and must not blindly believe them. So far, humans still possess superior intelligence and, most importantly, creativity that machines or algorithms cannot develop themselves.

There is one benefit of virtual conferences: Even if you missed it you can at least watch it again on YouTube.