Hannover Fair Industry HMI 2018

02/05/2018
  Copyright: © Image: Ahrens und Steinbach Projekte

“Artificial Intelligence in Production”: The Cluster of Excellence Production Technology and its network PROTECA (Production Technology Aachen) presented solutions all around the applicability of algorithms of machine learning for self-optimizing production technology at the HMI2018.

In 2018, the Cluster of Excellence “Integrative Production Technology for High-Wage Countries” of RWTH University and its network PROTECA were once again presenting their research at the “Hannover-Messe Industrie” (HMI). By now, the Cluster looks back on twelve successful years of top-level research in the field of production technology and looks purposefully forward to the future and the “Internet of Production” as the new core piece of Industrie 4.0. The role of PROTECA is to support the transfer of results and findings between industry and research institutions.

Within the framework of ProduktionNRW’s networkers PROTECA provided an interesting insight into recent Cluster research in cooperation with the Cybernetics Lab IMA/ZLW & IfU of RWTH Aachen and the Visual Computing Institute at the Hannover-Messe.

  Copyright: © Image: Ahrens und Steinbach Projekte

With the research project “CENSE” (“Cognition Enhanced Self-Optimization”) the applicability of algorithms in machine learning for self-optimizing production systems is investigated. The idea behind the research is based on the human learning process; the agent gains experience through numerous interactions with its surroundings. This enables it to autonomously learn an optimal behavioral strategy. Within CENSE, this approach is used in a real application scenario in the context of production by allowing a 6-axis industry robot to be autonomously controlled by the KI.

 

To achieve this the Cluster researchers of the Cybernetics Lab IMA/ZLW & IfU developed the “CENSE 2.0-Demonstrator” which is able to play the skill game “Hot Wire” by itself and demonstrates the agility and flexibility of path planning processes for stationary industry robots in consideration of external circumstances. The intelligent agent in the Cense-demonstrator is implemented by means of neural networks. Additionally, virtual reality glasses (VR) are used to depict these networks through 3D visualizations cast into the virtual room. The additional VR experience allows the visitors to understand the operating modes of artificial intelligences better.