Five minutes to go… 10 minutes to go… 20 minutes to go. Everyone who regularly waits for their train is familiar with this reverse countdown. Even in Germany, the country of punctuality, it is not unusual to bridge longer waiting times with podcasts. The statistics confirm the feeling of every train rider: in an EU comparison, Germany ranks third to last in terms of the punctuality of its trains. Deutsche Bahn is therefore turning to Industrie 4.0, using algorithms and artificial intelligence (AI) to tackle the problem. Especially with a complex system like rail traffic, planning is not easy. Hundreds of trains must be scheduled to the minute in a huge rail network. Delays and rescheduling of a train in Berlin can have effects as far away as Freiburg. In Stuttgart, Deutsche Bahn is therefore testing a system developed by the group’s own “House of AI”. An AI application determines the consequences of several decisions in the event of disruptions in the network. From this, a sort of video with recommendations for action is created. The dispatcher can fast-forward through the video and use it to make decisions.
Data-based decision support up to the takeover of decisions by AI are a pillar of the data-driven company of tomorrow. Not only the railroad but also every manufacturing company must ask itself how it can provide the user with optimal decision support in complex situations like planning, quality assurance or plant control. AI is therefore not a talking robot, but an algorithm in an application that, using trained data, calculates probabilities for events and provides the user with help in decision-making.
Manufacturing companies need to build significantly more competencies in the field through staff acquisition and a strong partner network. But available experts in the field are scarce and in demand. However, the right experts are not enough. The success factors of a sustainable implementation also include consideration of the company’s digital status quo: Where does my company stand on the path to Industrie 4.0? In which use cases is the greatest need and benefit? Because defining fancy AI use cases that do not consider the technical, organizational and cultural basis of the company leads to frustration and failure. Therefore, the current I.40 maturity of the plant should be assessed. The assessment focus must not only be on technical aspects but must also consider the organization (processes and workflows) and the culture (mindset and digital skills of the staff). In companies with multiple, sometimes globally distributed, plants, it is also necessary to set global standards that result from local needs and are implemented locally.
In our current issue of Quarterly Insights, we take a closer look at the application of artificial intelligence in manufacturing companies. Amongst others we have a podcast for you with Dr. Johannes Winter, CEO of Plattform Lernende Systeme – Germany’s AI platform, a book recommendation, and a special tip from the team for anyone who wants to understand how machine learning works in a practical way. Enjoy.