On 3 December 2018, a helicopter struck the earth wire of a high-voltage power line near Arbedo-Castione in Ticino, a few kilometres north of Bellinzona. In the causes section of its final report, the Swiss Transportation Safety Investigation Board, the federal agency that investigates aviation, rail and maritime accidents, writes that the accident is attributable to the fact “that the pilot did not perceive the obstacle owing to a restricted situation awareness”. The sentence that follows matters more: “The following factors, individually or in combination, may explain this restricted situation awareness.” Five of them follow. The reduced attention on a comparatively undemanding return flight to base. The obstacle situation in the departure direction, which the pilot did not call to mind before take-off. Misplaced priorities during departure. A collision warning device that was not installed. And crew resource management that never took hold. (Quotations from the Swiss reports are my translations from the German.)
Almost ten years earlier, in the final report on a collision with terrain near Aedermannsdorf in the Solothurn Jura, in north-western Switzerland, “insufficient situation awareness on the part of the pilot regarding the terrain and the tasks of air traffic control” appears as the third of four contributing factors. Nothing further stands beside it in the causes section. Anyone who pushes on into the analysis will indeed find the explanation: precision hand-flying in restricted visibility tied up a considerable share of the pilot’s cognitive resources, and radio contact with Basel Information fed the assumption that air traffic control was taking care of terrain clearance. Except that hardly anyone reads an accident report as far as the analysis. What gets read is the causes section, and what stands there is a label.
The same investigating body, operating in 2009 as the Aircraft Accident Investigation Bureau and since 2011 as the STSB. The same term, two uses. Which of them matches the model is best settled where the model comes from.
A book for people who build cockpits
Designing for Situation Awareness is not a book about investigation. It is a book about system design, and you can tell where it comes from. Endsley formulated her definition in 1988 as an engineer at Northrop, in the world of cockpit development, and worked it up into a theory in Human Factors in 1995. Situation awareness there is the perception of the elements in an environment within a given volume of time and space, the comprehension of their meaning, and the projection of their status in the near future. Hence the three levels that every human factors course now teaches: perceiving, comprehending, projecting.
The book she wrote with Debra Jones asks a prospective question: what must a system deliver so that all three levels can form in the first place? It answers with design principles for displays, alarms, automation and control rooms. The gain lies in the question not stopping at level 1. A display showing forty correct values and no trend serves perception and starves projection.
There is a measurement technique alongside it. SAGAT freezes a simulation at random points, blanks all the displays, questions the operator and compares her answers against the actual state of the simulation. A serious attempt to make a mental construct testable.
The figure everyone quotes
76.3 per cent. Given how often this number appears on training slides, it ought to be carved in stone. It says that 76.3 per cent of situation awareness errors in aviation sit at level 1, at plain failure to notice (Jones & Endsley 1996). Look into where it comes from and caution sets in. It rests on 143 voluntary reports to the American Aviation Safety Reporting System, retrieved by full-text search for the phrase “situational awareness”. The sample therefore consists of cases the reporters had already labelled as a situation awareness problem themselves. Add the bias in how people report: “I did not see it” is easier to admit than “I misread it”, which pushes level 1 upwards. The second figure, routinely fused with the first, is 88 per cent from NTSB accident reports, calculated only across accidents with a substantial human factors component, with 32 coded errors behind it. The direction holds all the same: most of what we call loss of awareness is failure to notice. It is just that this is an observation and not a measurement.
A report that names loss of situation awareness as the cause and leaves open how it came about has mistaken the name of the problem for its explanation.
What the book leaves open
The sharpest criticism appeared in 1995, in the very same issue as Endsley’s theory paper. As a description of a phenomenon John Flach considers the term useful; as a cause he considers it a circle. The event happened because situation awareness was lacking, and we know it was lacking because the event happened. A simple, easy to understand wrong answer, Flach writes, one that ends up standing in the way of research.
Twenty years later Sidney Dekker draws the practical consequence. Practitioners, he writes, now get accused of losing situation awareness in investigations, court cases and inquests, and as a human factors community we cannot walk away from our responsibility for having made that possible. The charge is not aimed at the model. It is aimed at the road the model has travelled: from a requirement placed on a display to an attribution placed on a person.
The third line of criticism is the most practical. Neville Stanton and colleagues describe situation awareness as a property of the socio-technical system, distributed across people and machines, whose shares need not be identical. Back to Castione: the flight assistant saw the power line from the beginning of the flight and recognised the developing collision early. But the crew were not connected to the on-board intercom at all and could communicate solely by radio. Between them stood a gap in experience that the STSB names explicitly, and a CRM course the flight assistant had completed online. The warning came immediately before the collision. Read individually, the pilot lacked the picture. Read as distributed, the picture was present in the system, and it had neither a channel nor permission to use one early.
Endsley pushed back in 2015 with a paper on seven misconceptions about her model. The three levels are ascending levels, she argues, not linear processing stages, and her model is not confined to the head. That is largely fair. Yet the reply does not close the decisive gap either: neither the book nor the defence says how the construct should be handled retrospectively. That is exactly the gap the reports fall into.
Who this book is for
For anyone who designs, procures or signs off displays, alarms, control rooms or automation. That is what it was written for, and for that it is good. If you investigate, take the model along as vocabulary for describing what someone was missing, not as a finding. The test fits on one line. If “situation awareness” appears in your report, do the conditions that constrained it appear in the same place? Not in the analysis on page eleven. Where the cause is stated.
Thoughts on this? Take it to LinkedIn.
No comments here – but discussion is welcome, where the readership is anyway.
Join the discussion on LinkedInSources
- Mica R. Endsley & Debra G. Jones – Designing for Situation Awareness: An Approach to User-Centered Design, 2nd edition, CRC Press 2012 (main source; 3rd edition 2025)
- Mica R. Endsley – Toward a Theory of Situation Awareness in Dynamic Systems, Human Factors 37(1), 1995, pp. 32–64
- Debra G. Jones & Mica R. Endsley – Sources of Situation Awareness Errors in Aviation, Aviation, Space, and Environmental Medicine 67(6), 1996, pp. 507–512
- Mica R. Endsley – A Taxonomy of Situation Awareness Errors, in: Fuller, Johnston & McDonald (eds.), Human Factors in Aviation Operations, Avebury Aviation 1995, pp. 287–292
- John M. Flach – Situation Awareness: Proceed with Caution, Human Factors 37(1), 1995, pp. 149–157
- Sidney W. A. Dekker – The Danger of Losing Situation Awareness, Cognition, Technology & Work 17(2), 2015, pp. 159–161
- Mica R. Endsley – Situation Awareness Misconceptions and Misunderstandings, Journal of Cognitive Engineering and Decision Making 9(1), 2015, pp. 4–32
- Neville A. Stanton et al. – Distributed Situation Awareness in Dynamic Systems, Ergonomics 49(12–13), 2006, pp. 1288–1311
- STSB (SUST) – Final Report No. 2405, helicopter HB-ZCM, Arbedo-Castione, 3 December 2018
- Aircraft Accident Investigation Bureau (BFU) – Final Report No. 2111, aircraft N401AC, Aedermannsdorf, 8 February 2009