Concept For Automated Economic Evaluation Of AI Applications In The Production

An unclear economic benefit is one of the biggest obstacles for companies to use AI applications in their production. Various methods exist for deriving the economic efficiency of AI applications in a company's own production, but these have to be considered individually for each use case. In addition, they rely on the statements of the software manufacturers with regard to efficiency gains, quality gains, and other KPIs. In the course of the FabOS research project, a concept was developed that uses Asset Administration Shells and ML clustering for the automated economic evaluation of AI applications in the production.