Künstliche Intelligenz

To manage the increasing complexity in current hardware design processes, current systems are increasingly designed on abstract layers. While the more rapid development of prototypes is a clear advantage of this paradigm, these designs suffer from being closed up and hard to analyze. There is no simple way to extract a system’s structure from its description anymore. Nevertheless, the designers should get all the information they need during development. The computer is assisting in this process with the observation of its inner self: The simulated hardware is supervised by an artificial intelligence (AI). It learns about a system’s functions while the system itself is running. Dependencies and connections inside this system are retrieved independent from their availability, thus speeding up the development process.

As a result of continuously increasing economic constraints for the producing sector in Germany due to competitors from low-wage countries, production paradigms are changing substantially. This paradigm shift is characterized by rapidly advancing automation processes. Hence, highly complex Cyber-Physical Production Systems (CPPS) are developed and put into practice. In combination with decreasing numbers of staff and the resulting loss of knowledge, this leads to a deficit of competence required to handle the increasing complexity of CPPS. As a result, a need for innovative assistance systems arises for the support of the remaining employees. The given article describes the challenges and problems and drafts a potential technical solution as well as challenges concerning organization and staff.

Today sensors and actuators gain increased importance in industrial production processes. Traditionally centralized processing architectures are used for sensor data processing and actuator control. Increasing sensor and actuator densities require new decentralized and distributed processing methods and architectures. Artificial Intelligence, part of computer science, can contribute to robustness and autonomy aspects concerning data processing and distribution.

In the planning phase of cyclically interlinked production lines neural networks are used to learn and forecast the characteristics of these systems. To increase the results of learning and to accelerate the training this paper presents a method, based on genetic algorithms, that reduces the attributes to describe the behaviour of the production lines.