A strong mix of research results and industry presentations was presented by the Digitalization and Information Management program of the Stuttgart University of Applied Sciences (HFT) on October 19. The 120 or so guests - students, alumni and representatives from industry and business - also had plenty of opportunity to network.
Advanced sensor technology, efficiency gains through AI and cleverly programmed software solutions in the control & monitoring of machines, logistics and industrial plants as well as the topic of autonomous driving were the focus of the event, which was moderated by Prof. Dr.-Ing Detlef Pape, Prof. Dr.-Ing. Dieter Uckelmann and Robert Otto. Prof. Dr. Katja Rade, Rector of the HFT, welcomed the participants with a convincing AI example.
Fast and flexible
Martin Stühmer and Nima Samsami (CGI Germany) explained how they are already using prompt engineering in their software programming and testing to achieve the desired results around 15% faster and thus much more efficiently. What is important here are competent software developers who can recognize and correct the errors that the AI produces (hallucination). "I haven't analyzed any code for errors on my own for a year now," says Martin Stühmer with a smile. A live demonstration and tips for prompting rounded out the talk.
Oliver Ruzek, CEO of NC-Vision explained the use of 'no-code technologies' in digitalization in production. Today, he said, flexibility and resilience are even more important to the industry than pure cost savings. According to Ruzek, monolithic software solutions with hundreds of thousands of lines of code, while widely used, are the wrong approach to changing requirements. Using the platform developed by NC-Vision, application software can be assembled 5-10 times faster with a flexible toolbox of pre-programmed and reusable software components. "We help manufacturing companies master the digital transformation" said Ruzek.
Rüdiger Thomas Kreis (ICS Group) explained the hybrid ICS complete solutions using the example of optimized logistics. "Our customers always have 100% control," Kreis emphasized. The solutions are based on a digital twin for the goods, the handling unit and the warehouse, as well as a combination of suitable hardware (CVis/Scangates), some of which was developed in-house, and software created specifically for the requirements - including suitable interfaces.
Kerstin Bergmann (Bosch) described the challenges in the area of sensor technology in autonomous driving. "The vehicle must be able to do everything a human can do," Bergmann noted. Particularly at SAE levels 4 and 5, where the driver 'switches off the brain' or is only a 'passenger', the challenges are enormous. In the meantime, the combination of many specialized video, radar and ultrasonic sensors is generating an enormous amount of data. This has to be processed rapidly so that the vehicle can constantly recognize the environment and the situation and derive the right actions from it. For this purpose, the sensor technology is usually separated from the further processing. At Bosch, hardware and software are also developed separately and modularly for the areas of assistance, mobility, comfort, energy, and infotainment.
Current, practical projects of the research group Digitization and Information Management were presented by the team around Prof. Uckelmann and Prof. Pape: Ferdinand Munz presented the project KNIGHT (AI-supported learning & teaching), Sergej Kreber the project SenSim4iCity / People Counting, Robert Otto the project UDigiT4iCity TP2 / Energy Monitoring, Salam Traboulsi the project UDigiT4iCity / 5G Networking and Andreas Trinkle the project SenSim4iCity / TP 1, which deals with automated analysis of building control systems. In addition, Antonino Forte presented his master's thesis. In it, the HFT graduate investigated whether consumer-driven contracts bring an advantage in the delivery of software, what requirements a system must fulfill in order to be able to use them, and what alternatives there are. Afterwards, the participants continued their discussions in block 4.