Some Issues with Interaction Design and Implementation in the Context of Autonomous Interactive Critical Systems

Increasing analysis and decision capabilities in systems is a classical approach followed by designers and engineers to support operators in the command and control tasks of more and more complex systems. However, designing interfaces that will be used to operate intelligent systems and (partly) autonomous systems is a very complex activity altering in depth the development process of these systems. In the early days, the development of such systems was rather limited and thus easier to manage even though bad designs have been widely and frequently reported (e.g. [13, 15]). Nowadays, they can be integrated to several critical systems, such as Air Traffic Control management systems, as illustrated by the example presented in the case study. This position paper raises some issues related to the design and implementation of safety-critical user interfaces featuring intelligent and (partly) autonomous behaviors.