VAPi-KI
Verfication process for requirements in the processindustry based on huge and unstructured data with artificial intelligence
The Distr@l project started on May 1, 2024, the scheduled ending is on October 31, 2026. The financing is partly realized by funding from the development of innovative digital processes and is supported by the Hessian Ministry for Digitalization and Innovation.
For the project, the highly specialized software company Conweaver GmbH and the renowned pmd (product development and machine elements) department of the Technical University of Darmstadt, led by ConSenses, are forming a project consortium to tackle a challenging task. Werner Schmid GmbH, Andritz Kaiser GmbH and :em AG have been brought on board as associated partners. Together, the partners want to bridge the gap between the development and operating phases of industrial goods in the process industry. The project is being developed and tested using specific examples of punching and forming machines. These systems are used in series production of a wide variety of metal products (from electronic contacts to fittings and vehicle components, coins or general consumer goods) and are subjected to a wide range of often unknown loads over decades under a wide variety of operating conditions, which often cannot be anticipated in the design phase of subsequent systems or during repairs, maintenance or adjustments properly.
For practical grounding and testing, machine strokes of a system have been comprehensively recorded and evaluated at the associated partner Werner Schmid in Fulda since the first month of the project. At the same time, the machine manufacturer Andritz Kaiser is disclosing key design principles and details that form the basis of the machine design. The project partners Conweaver, pmd and ConSenses want to structure, make accessible and connect this diverse, sometimes highly unstructured information using the graph technology Linkshpere from Conweaver, Natural Language Processing (NLP) from pmd and clustering algorithms from ConSenses, so that systematic lessons can be learned from daily machine loads for future machine designs. Please contact us or project partners if you would like further information.