Carl Westin

92 Automation transparency effects Voshel demonstrated how an EID-derived interface for the supervisory control of unmanned vehicles in the maritime domain can be used to increase the transparency of system decision-making and behavior. 227 Similarly, and as noted earlier, Chen and colleagues showed that merging EID with the SAT framework can benefit sys- tem trust and situation awareness, without increasing workload. 200, 215, 216 The approach centers around a work domain analysis of the sociotechnical sys- tem considered, 122 which typically consists of decomposing the domain using an abstraction hierarchy. 228 This hierarchy describes the system at different levels, ranging from the functional high-level purposes of the system to more specific gen- eralized functions and the physical form of system elements and components, which serves to determine which information to visualize. In order to determine how this information should be visualized, the information requirements are considered in relation to human information processing abilities and problem-solving behaviors using, for example, the skills, rules, and knowledge (SRK) framework. 85 Because EID visualizes a complex system across multiple levels of abstraction, it provides a useful scaffold for supporting deep, knowledge-based reasoning over system be- havior during novel situations or fault response. 122 5-3 Ecological displays in ATC CD&R Importantly, the EID paradigm is believed to correspond well with how skilled op- erators perceive and solve problems, namely by considering the relationships of ob- jects and events in the world, rather than the precise stimulus elements. Similarly, it has been suggested that air traffic controllers base their judgments and decisions in CD&R on the relationships and interactions between aircraft and their constraints as they evolve over time, rather than on information about the aircraft or its po- sition. 229 However, these relationships, and the possibilities for solving conflicts that they afford, are not readily accessible to controllers in current ATC systems. Rather, controllers compute these relationships by engaging in effort and time de- manding inferential cognitive processes, interpolating multiple information sources. In order to balance the cognitive load in CD&R, research suggests that controllers rely heavily on decision-making heuristics. 79, 80 An example is the use of trajectory prediction strategies that involve mental extrapolation of aircraft trajectories in con- flict detection. Strategies include time comparisons, distance comparisons, altitude comparisons, and contraction rates. 26, 33 Over the last 30 years, several ATC research projects have explored interface de- signs that to some extent acknowledge the work ecology, and attempt to alleviate the cumbersome extrapolation involved in CD&R. These interface typically visualize obstacles (e.g., other traffic, restricted areas, weather, terrain) as constrained “no-