Safety, Aviation and Human Factors
Lessons learned from accident reviews
by Jose Luis Garcia
As you all know, safety, aviation and human factors are closely related. I am not revealing anything new there. However, the review of any accident provides good material for reflection. I hope you enjoy reviewing the following cases as much as I have. Many of them engaged me directly with my target career, human factors and aviation. Perhaps you will find them relevant as well.
Historically, 70 percent of commercial jet accidents involved humans (Boeing 2004), in one form or another. In the early years of aviation, technical defects were the main cause of accidents. But as reliability improved, the human factor became the main factor.
Some say human error is the primary factor in up to 80 to 90 percent of the cases. However, aviation accidents often involve more than one cause, such as system complexity, poor human-machine interface, inappropriate work organization, awkward work procedures, altered communication between pilot and controllers, or loss of situational awareness. We should not place blame on the human operator alone, but rather to a combination of factors in a highly complex context.
The best way to understand the human factors of the aviation world, and how they can be applied, is to review some of the big accidents that have occurred. We will find there the “Aha!” The analysis of two of these accidents motivated me to pursue deeper human factors education: the accident at Los Rodeos airport in Tenerife (Spain) in 1977, where two B747 (from KLM and Pan Am) collided on the runway; and the accident of Cali (Colombia) in 1995, where an American Airlines B757 crashed into the terrain. Each aviation accident provides myriad lessons to learn, but those two are especially relevant, in my opinion.
LOS RODEOS RUNWAY COLLISION
The Los Rodeos runway collision remains the worst accident in aviation history, hopefully forever, with 583 fatalities. This accident was a clear example of the relevance of human factors in aviation. Stress, decision-making errors, communication problems between pilot, controller and crew, lack of situational awareness, and organizational management issues in the cockpit were contributing factors.
Los Rodeos runway collision record:
http://aviation-safety.net/database/record.php?id=19770327-1
Los Rodeos accident report:
http://ntl.bts.gov/lib/7000/7500/7585/jatww3-1wilson.pdf
CALI B757 TERRAIN CRASH
The Cali B757 terrain crash was the first accident of a B757. By that time, it was the only airliner that had not had fatal accidents, along with the Concorde. It made clear how human interaction with automated systems in aviation required further improvements, and provided new design opportunities of human-computer collaboration. The accident in Cali had the following contributing factors: lack of situational awareness due to poor information, interaction problems with automated systems, human dependency on automation, work load when manipulating automated systems, and usability issues of critical systems.
Cali B757 terrain crash record:
http://aviation-safety.net/database/record.php?id=19951220-1
Cali B757 accident report:
http://www.cs.bath.ac.uk/~anneb/collwn.pdf
GERMAN MIDAIR COLLISION
But if those examples are not enough, we could remember the midair collision in Ueberlingen, Germany in 2002. A DHL cargo B757 collided with a Bashkirian Airlines Tupolev 154 at 35,000 feet. They were the only two aircraft in the controlled sector by Zurich ACC. How could this be possible? This time the controller was not overloaded, and the pilot did not misunderstand his instructions. However, problems with divided attention, high degree of trust in automation systems, macroergonomic issues (shift organization), differences on training due to cultural differences, and decision-making in critical situations were the decisive factors.
German midair collision analysis:
http://www.aviation.uiuc.edu/UnitsHFD/conference/humfac04/nuneslaur.pdf
IT’S NOT ALL ABOUT ACCIDENTS
Similarly, by reviewing other accidents, we will see other areas to apply human factors. But human factors isn’t only about preventing accidents. According to International Civil Organization Association (ICAO), human factors in aviation hopes to help achieve the aviation system goals of safe and efficient transportation of passengers and cargo. Efficiency should not be forgotten, and again, the human being is the key.
To visualize human factors concepts, ICAO uses the SHELL model as introduced by Edwards in 1972, and modified by Hawkins in 1984. SHELL stands for Software, Hardware, Environment, and Liveware.
The approach to aviation systems design, and in general to any system, should be comprehensive. It should integrate technical, human and procedural systems to deliver the desired system performance. The result will be a system that reinforces human capabilities, mitigates human limitations, and fulfills human needs.
Human factors will be always an issue in aviation, as long as pilots control aircraft, and air-traffic controllers are in charge of aircraft separation. Even in the case of unmanned vehicles, in free-flight airspace, human operators will monitor the traffic situation and intervene for maintenance. Human factors will be there. There are so many things to understand, so many things to solve, and so many designs to improve. We have such a huge field to explore!
To get a start, explore these links:
EUROCONTROL HF
http://www.eurocontrol.int/humanfactors/public/subsite_homepage/homepage.html
FAA human Factors Division
http://www.hf.faa.gov
NASA HF research and technology division
http://human-factors.arc.nasa.gov
Flight Safety Foundation
http://www.flightsafety.org
CAP 719 Fundamental Human Factors Concepts (UK CAA 2002)
http://www.cs.bath.ac.uk/~anneb/collwn.pdf
CAP 737 CRM (Crew Resource management) Training (UK CAA 2003)
http://www.caa.co.uk/docs/33/CAP737.pdf
