Advanced technologies and cockpit design. Have we learned the lessons of the past?

In the 1980s, the development of automated systems in the cockpits began to take shape. While it was clear that automation would play a more prominent role in aviation, its impact on crew performance lacked substantial research support (Wiener, 1989; Wiener & Curry, 1980).

As automated systems advanced in the 1990s, new theoretical research came up on automation use (Parasuraman & Riley, 1997; Parasuraman et al., 2000), mode awareness (Sarter & Woods, 1995), and the startle and surprise effect (Sarter et al., 1997). However, if, on the one hand, these studies supported a thoughtful inclusion of technology, highlighting that, in addition to the undeniable benefits, new problems would emerge, on the other hand, they favoured the propensity for a "rush to implement" (Salas & Maurino, 2010, p. 7) automation.

On the 26th of April 1994, China Airlines flight 140 crashed while approaching Nagoya airport in Japan after the crew attempted to recover the aircraft from the consequences of inadvertently activating Go/Around mode by the First Officer.

Two hundred sixty-four people died (including two infants and 15 crew members), and seven passengers were seriously injured as a result of this accident.

The final report of the Japanese Air Accident Investigation Commission (AAIC), published in 1996, highlighted some critical points: insufficient crew understanding of the operation of onboard automatic systems and deficiencies in the design of the computerised flight control system of the aircraft. AAIC, therefore, issued many recommendations.

China Airlines was required to improve training to ensure that pilots fully comprehend the Auto-Flight System (AFS) functions on advanced-technology aircraft (AAIC, 1996).

The two involved authorities–the French and the Taiwanese civil authorities–were requested to reinforce education and training on the design concept of the advanced technologies. The aircraft's Auto Flight System (AFS) was too complicated to interpret and manage in certain conditions: "There are occasions where it would be difficult for pilots to recognise the operating conditions of the AFS or properly predict the effect of a mode change on the flight. There is the possibility that the crew might be unable to take proper action when using functions rarely used on daily flights. Because human thinking ability is restricted in times of high stress […], it would be even more difficult for crews to take action within a limited period […] The design of the AFS (function, mode display method, warning and crew recognition function) should be reviewed, taking into account the pilot's behaviour and human cognitive process under an emergency or abnormal situation." (AAIC, 1996, p.104)

Airbus Industrie was requested to improve the design of some of the cockpit's advanced functions: the disconnection and override of the Autopilot (AP) and the warnings and recognition functions for Trimmable Horizontal Stabiliser (THS) movement (AAIC, 1996).

After the report's publication, the Federal Aviation Administration (FAA), in collaboration with the European Joint Aviation Authorities (JAA), commissioned a study to issue a list of recommendations "for the future design and certification of pilot interfaces. automation." The study once remarked on the close correlation between advanced technology and automation of modern aircraft and the human factor (Demagalski et al., 2002).

In DOC 9995 of 2013, the International Civil Aviation Organization (ICAO) acknowledged that more needed to be done in the past to address the complications arising from the evolution of modern aircraft design and the environment (ICAO, 2013).

Over the past decade, the rapid advancement of intelligent technologies has generated both enthusiasm and increasing apprehension. The most sophisticated models exhibit significant power but are unpredictable, opaque, and prone to errors. Researchers and regulators are concerned about this "rush to implement" (Salas & Maurino, 2010, p. 7) Artificial Intelligence (AI), where companies or states may lower safety standards to outpace their competitors in innovation.

In a previous study titled "The limits of data-driven technologies in the world of complexity", I explored the limits of these technologies in the complexity of today's systems. I pointed out how contemporaneity is characterised by unprecedented events that cannot be predicted a priori by any intelligent technology, as these technologies are based on past events. Consequently, humans remain essential and central to contemporary systems, playing a crucial role in making final decisions and taking action.

Therefore, integrating advanced technologies into the cockpit must recognise pilots' contributions. Their involvement is crucial for flight operations' safety and efficiency. The more active they are in these challenging changes, the more confident they will use these advancements.

The risk is that, on the wave of general enthusiasm, someone designs and introduces technologies that operators have not fully understood, unsuitable and complex to manage when operating conditions are "limited to periods of high stress [when it is] more difficult for crews to intervene in a limited period" (AAIC, 1996, p.104).

The consequences will be severe and intolerable, as we had the responsibility to learn from the past this time.

References

AAIC (1996). Aircraft Accident Investigation Report. China Airlines, Airbus Industrie A300B4-622R, B1819. Japan Aircraft Accident Investigation Commission. Nagoya Airport. AICC. Japan.

Demagalski, J., Harris, D., Salmon, P.M., Stanton, N.A. (2002). Design Induced errors on the modern flight deck during approach and landing. HCI-Aero: Human-Computer Interaction in Aerospace conference. Boston, USA.

Hunt, W (2020). The Flight to Safety-Critical AI. Lessons In Ai Safety from The Aviation Industry. UC Berkley.

ICAO (ed.) (2013) DOC. 9995, Manual of Evidence-based Training. First Edition. International Civil Aviation Organization, Montreal, Quebec, Canada.

Parasuraman, R., & Riley, V. (1997). Humans and automation: Use, misuse, disuse, abuse. Human Factors, 39 (2), 230 – 253.

Parasuraman, R., Sheridan, T. B., Wickens, C. D. (2000). A model for types and levels of human interaction with automation. IEEE Transactions on Systems, Man, and Cybernetics-Part A: Systems and Humans, 30 (3), 286–297.

Salas, E., Maurino, D. (2010) Human Factors in Aviation, 2nd edition. Elsevier.

Sarter, N., & Woods, D. (1995). How in the world did we ever get into that mode? Mode error and awareness in supervisory control: Situation awareness. Human Factors, 37 (1), 5–19.

Sarter, N., Woods, D., & Billings, C. (1997). Automation surprises. In G. Salvendy (Ed.), Handbook of human factors/ergonomics (2nd ed.) (pp.1926–1943). New York: Wiley.

Wiener, E. L. and Curry, R. E. (1980), Flight Deck Automation: Promises and Problems. Academic Press01.

Wiener, E. L., Curry, R. E. (1980) Flight Deck Automation: Promises and Problems. NASA, Washington, DC, United States. NASA, Washington, DC, United States.

Wiener, E. L. (1989). Human factors of advanced technology (“glass cockpit”) transport aircraft (NASA Contractor Rep. No. 177528). Moffett Field, CA: NASA Ames Research Center.