The development of intelligent systems that collaborate with humans requires multi-disciplinary research with expertise from computer science, AI, and social sciences. However, central to this process are experts and users from the domain in which the system will operate. By understanding users and by collaborating with them in an iterative manner, one can ensure that the system and its functionalities add value to the current work context.
Our HART-team has developed a methodology to perform iterative, incremental research and development for human-agent system design (known as “Socio-cognitive engineering”). We apply this method for example in the domain of healthcare, in which we investigate how clinical decision-support systems can be used by clinicians to facilitate the diagnostic process. By involving domain experts in all stages of our research, we gain valuable insights with respect to the operational and sociological demands for such systems.