Limitations of Activity-Centered Design for Air Traffic Control Systems

Human-centered design (HCD) and activity-centered design (ACD) are two approaches to designing a system or station. They each have their strengths. For example, HCD considers human limitations while ACD considers the task goals that need to be accomplished. An ideal design would pass both HCD and ACD design checklists. However, if one approach has to be dominant, there are more observable downsides to an ACD design. When designing a system for air traffic controllers (ATCs), if moving from an HCD to ACD approach, the design could sacrifice human limitations both mentally and physically, introducing high chances of human error that affect the outcomes and performance. Additionally, using the HCD approach, designers, engineers, and scientists can create a driving force towards advanced technology that enhances human-machine interaction. 

Mental limitations

The work of an ATC involves complex cognitive activities, including storing information in working memory, analyzing data, and communicating with pilots that largely depend on human capabilities (Inoue et al., 2012). According to Miller’s law, the human working memory has a capacity of seven plus or minus two chunks (Guastello, 2013), suggesting there is a limit to human cognition. With the increase in travel demands, ATC’s workload has also become heavier (Inoue et al., 2012). When an ATC system is an ACD, since it is activity-based, it would be deliberately calculated how many aircrafts each ATC would need to communicate with simultaneously and design the system based on the calculations. However, without thinking about cognitive workload and mental capacity, the system could induce an intolerable load on an operator’s memory. When working memory is experiencing a heavy load, the central executive lacks the additional attention span to other cognitive tasks such as noticing important stimuli and notifications. As Inoue et al. (2012) stated, errors often occur when the mutually dependent relationship between humans and machines breaks down. When operators can no longer manage the workload given by the system, errors occur. 

Physical limitations

ATCs' work involves managing multiple large screens and identifying visual cues while performing other cognitive tasks. The notifications and visual stimuli become important information being transferred in this multitasking environment, and monitoring this information is an essential part of the task (Imbert et al., 2014). Because there are multiple screens, it is impossible to have all the notifications at the center of the visual field because the system needs to make sure the visual stimuli do not interfere with the main task. Therefore, ATCs utilize peripheral vision in the visual perceptual system to perceive any notifications or change in aircraft position in radars for farther away screens at the workstation (Imbert et al., 2014). This raises the issue of differences in anthropometry and individual differences. For example, ACD shows the system requires the use of a three by three screen setup with a total of nine screens to produce the most efficient workflow. 

Individuals who are shorter or with smaller areas of peripheral vision will not see changes in the screen. They will need to move their seats to see the side screens' outer area, which decreases efficiency or completely omits the notifications. Because peripheral items have reduced salience (Imbert et al., 2014), it is even more critical to make sure it is within an individual's visual field. Therefore, without considering individual differences in anthropometry and merely relying on the engineered formulas and calculations to set up the most optimal workstation, human operators cannot utilize the workstation to the fullest and even introduce omission errors.

Drive for technological advancement

Moving from an HCD to an ACD approach runs the risk of slowing down technological advancement. Many new technologies are designed with human users in mind to close the gap between humans and machines by minimizing human limitations, allowing machines to serve at their fullest potential. For example, Tran Luciani et al. (2019) experimented with a fine-grained automation system for ATC by sketching on duty, improving operators’ reflection time, and holding less in their working memory. One of Norman’s (2015) arguments for ACD is that the HCD can result in a design that is too complex because designers are listening and improving on all complaints made by operators. However, as William (2009) comments, HCD is not about directly asking what users want, but to use a human factors professional’s expertise to profile users, define their behavior, and make design decisions based on that information.

Conclusion

Ideally, a design takes into account both the activity and the human aspects. However, if moving further away from HCD to focus on ACD, the design would run into risks such as sacrificing human mental and physical limitations and individual differences. In the case of ATC, where the work involves heavy cognitive load and visual stimuli detection accuracy, errors can form when a design does not consider major principles of HCD. Furthermore, because of the identified human limitations, technology advances to compensate for human limitations. Although one could argue, technology invention with the ACD approach could also lead to technological advancement, without considering human limitations, fewer users will be able to use those technologies as intended. Therefore, when designing, it is still best to consider the human-centered approach.

References

Guastello, S. J. (2013). Human Factors Engineering and Ergonomics: A Systems Approach, Second Edition (2nd ed.). CRC Press.

Imbert, J.-P., Hodgetts, H. M., Parise, R., Vachon, F., Dehais, F., & Tremblay, S. (2014). Attentional costs and failures in air traffic control notifications. Ergonomics, 57(12), 1817–1832. https://doi.org/10.1080/00140139.2014.952680

Inoue, S., Furuta, K., Nakata, K., Kanno, T., Aoyama, H., & Brown, M. (2012). Cognitive process modelling of controllers in en route air traffic control. Ergonomics, 55(4), 450–464. https://doi.org/10.1080/00140139.2011.647093

Norman, D. A. (2005). Human-centered design considered harmful. Interactions, 12(4), 14–19. https://doi.org/10.1145/1070960.1070976

Tran Luciani, D., Löwgren, J., & Lundberg, J. (2019). Designing fine-grained interactions for automation in air traffic control. Cognition, Technology & Work, 22(4), 685–701. https://doi.org/10.1007/s10111-019-00598-9

Williams, A. (2009). User-centered design, activity-centered design, and goal-directed design. Proceedings of the 27th ACM International Conference on Design of Communication - SIGDOC ’09, 1–8. https://doi.org/10.1145/1621995.1621997