AI for Predictive Costing: Artificial Intelligence Improves Efficiency and Accelerates Innovation Processes in Product Costing
In early June, twenty experts from industry and research discussed the use of artificial intelligence in product costing. The host of the roundtable event was FACTON, a company that recently launched the “AI for Predictive Costing” initiative to define meaningful ways to use AI in the field of costing. Participants talked about challenges and their vision of the future. Microsoft and the German Research Center for Artificial Intelligence (DFKI) were knowledge partners of the event.
FACTON, the company that manufactures FACTON EPC product costing software, launched the “AI for Predictive Costing” initiative to develop various scenarios to-gether with experts from industry and research using big data and artificial intelli-gence in product costing. The executive roundtable in the historic “Weisse Villa”, which used to be known as “Villa Metz” in Potsdam, kicked off the event.
“A constantly changing market environment and new technologies not only allow us to, but also demand that we take a new view of product costing. The objective of this kick-off event was to discuss which requirements artificial intelligence will need to satisfy in order to become relevant in this context,” says Alexander M. Swoboda, CEO at FACTON, describing the event’s background. “The discussion shows that not only will we be able to use AI to complete existing tasks more efficiently, we will also uncover entirely new solution paths that until now had been unfeasible or even inconceivable”.
Application Scenarios Featuring Artificial Intelligence in Product Costing Today there are already three models that demonstrate options for using artificial intelligence in product costing:
An automated cost estimate for the early stage of product development based on a similarity matrix using historical customer data.
An outlier analysis based on customer data that enables businesses to highly standardize cost estimates.
Automated costing of parts using algorithms – building on features of CAD models and components that have already been manufactured.
Alongside these concrete application possibilities, participants also discussed vision-ary ideas relating to the topic of “AI for Predictive Costing”. For example, visualizing current information on machine, wages and energy costs is now conceivable thanks to the integration of real-time data. An AI system can automatically detect relevant changes. In the future, AI assistants will take over routine tasks such as researching and inputting the necessary data in the area of Should Costing. A highly interesting topic in the area of cost control is the intelligent risk assessment of procurement processes, material prices and wages.
Current Challenges To use artificial intelligence as an innovation accelerator and exit the experimentation phase, acceptance of the technology must be increased. This will mean focusing more on the topic of “Explainable AI”. After all, to accept an AI decision, people need to understand the reasoning behind it. The very real technical gap between “can” and “want” must also be bridged. The first step is adapting existing IT infrastructures to the sophisticated demands of new IT services, processes and security mechanisms. Finally, the legal and regulatory environment must be studied carefully. Issues such as data protection and data privacy, intellectual property protections and product liability are key levers in making the use of intelligent systems ubiquitous. AI models will never fully take hold in industry without a competitive legal framework and a harmonized policy.
Note to editors: We are happy to provide additional information on “AI for Predictive Costing”. Mr. Alexander M. Swoboda is available for interviews.