A Conceptual Framework for Evaluating the Transferability of Production Management Decisions to AI

The field of artificial intelligence (AI) has advanced rapidly in recent years, demonstrating remarkable capabilities in solving complex decision-making tasks once considered exclusive to human intelligence. The potential of AI has been showcased through its success in strategic games, where AI systems have already surpassed human performance in structured, rule-based environments. This achievement raises the fundamental question of whether similar decision-making abilities can be applied to production management (PM) context that involves uncertainty, dynamic interactions, and strategic trade-offs. However, in the field of PM, the systematic application of AI for decision-making is still emerging and no comprehensive and controlled approach has been established to systematically compare human and AI performance. The research project "Manager ex Machina – Investigation of the Transferability of Production Management Decisions to AI Systems" aims to develop a simulation-based method that enables this evaluation and comparison of decision-making performance. Therefore, a simulation will be developed as a serious game that enables both human participants and AI agents to make PM decisions within a controlled and reproducible environment. The decision scenarios forming the serious game will be derived from expert interviews with specialists in PM and AI to identify decisions with the highest automation potential. The resulting data from the serious game will then be analyzed to evaluate the effectiveness and limitations of AI performance in these specific PM decisions. This allows to establish a conceptual framework for evaluating AI decision-making performance that can be transferable to other managerial domains, including workforce planning and supply-chain management. Future studies should focus on validating and refining the proposed concept through empirical studies and the development of a serious game. This will contribute to the development of a systematic benchmarking framework for evaluating human – AI performance in decision-making contexts.