adaptivity

Designing adaptations in AI-based work systems poses a central challenge for achieving human-centered AI (HCAI). This paper presents a methodical approach that enables transdisciplinary teams to systematically explore and structure the design space of adaptable work systems. Building on an extended work system model and operationalized through a matrix-based framework, the method supports the identification of interdependencies, stakeholder perspectives, and context-specific goals. Its practical applicability is demonstrated through a real-world case study in radiographic non-destructive testing.

Adaptation of work to humans instead of humans constantly adapting to work systems and work environments has been a goal of ergonomics since the 19th century. To date, an actual adaptation of work to the specific features, conditions and requirements of individual users has so far only taken place in rudimentary form. The possibilities have now opened up by the advances of Cyber Physical Production Systems (CPPS). This enables a renewed attempt to implement the goal of adapting work systems to workers. Integration of advanced sensory skills into work systems within manufacturing allows better recognition of the work environment and the actual state of the system. Together with actuator capabilities, it enables adaptive work systems to adjust their structure and behaviour to the anthropometric and the cognitive features and requirements of the systems’ users. However, a framework that integrates the concepts of adaptivity and personalization into the work system model and work system ...