Some pieces out of the final report from Gulf Air manamana 2000
final report
The accident itself has nothing in common with AF447, but it highlight the human factor somewhat closer. That might help the discussion in understanding the difference between blame and detailed accident investigation.
2.3 Analytical Methodology
A review of the factual information indicates that this accident was primarily
attributable to human factors, there being no technical deficiencies found with the aircraft and its systems. Consequently, the following analysis focuses on these human factors issues, both at the personal and the systemic levels. The analysis adopts the philosophy of Annex 13, which is well articulated by Dan Maurino, Coordinator of the Flight Safety and Human Factors Study Programme, ICAO. ‘To achieve progress in air safety investigation, every accident and incident, no matter how minor, must be considered as a failure of the system and not simply as the failure of a person, or people’.
The term ‘human factors’ refers to the study of humans as components of
complex systems made up of people and technology. These are often called ‘sociotechnical’ systems. The study of human factors is concerned with understanding the performance capabilities and limitations of the individual human operator, as well as the collective role of all the people in the system, which contribute to its output. There are two primary dimensions of human factors, these being the individual and the system.
In this context the following analysis addresses the human factors issues: at
the individual level, and at the systemic organisational and management level.
2.3.1 Individual Human Factors
In considering the role and performance of individuals it must be recognised
that people are not autonomous, they are components of a system. Therefore
human performance, including human errors and violations, must be onsidered in the context of the total system of which the person is a part. There is a need to investigate whether such errors or violations were totally or partially the products of systemic factors. Some examples are: training deficiencies, inadequate procedures, faulty documentation, lack of currency, poor equipment design, poor supervision, a company’s failure to take action on previous violations, commercial pressures to take short cuts, and so on.
2.3.3 The Reason Model of Safety Systems
At the 1992 ICAO AIG meeting it was recommended that the Reason Model
should be used as a guide to the investigation of organisational and management factors.
The Reason Model is described in the ICAO Human Factors Training
Manual (1998, Chapter 2). The model and its application is described in more detail in the book Managing the Risks of the Organisational Accident (Reason, 1997).
Operational experience, research and accident investigation have shown that
human error is inevitable. Error is a normal characteristic of human performance and while error can be reduced through measures such as intensive training, it can never be completely eliminated. Consequently, systems must be designed to manage human error. What follows is an integrated systemic analysis based on information drawn from all the specialist groups involved in the investigation. It is conceptually based on the Reason Model of safety systems.
2.4.6 Information Overload
The circumstances in the cockpit, and the behaviour of the captain, indicated
that at this time (1929:41) the captain was probably experiencing information
overload. While there are a number of theories of human information processing, one characteristic that they all share is the concept of some form of overall central limitation on the rate at which humans can process information. This may take the form of a ‘bottleneck’, a pool of limited attentional resources, or an ‘executive controller’, supervising and co-ordinating multiple information processing resources.
However, while the underlying more esoteric theoretical issues continue to be investigated, the research carried out over the last 50 years or so, combined with actual operational experience has provided a practical first order working model of the fundamental capabilities and limitations of human information processing. This model is applicable to ‘real world’ situations, such as the analysis of human performance in complex socio-technical systems, accident investigation and training.
Some key aspects of the model are briefly described as follows:
At the conscious level, the human brain functions as if it were a single channel information processor of limited capacity. Under conditions of information overload, responses fall into one or more of the following categories:
Omission - ignore some signals or responsibilities.
Error - process information incorrectly.
Queuing - delay responses during peak loads; catch up during lulls.
Filtering - systematic omission of certain categories of information according to some priority scheme. This can lead to the focussing, or ‘channelling’ of conscious attention on one element of a task, or situation, to the exclusion of all others.
Regression - reversion to a previously over-learned response pattern.
Approximation - make a less precise response.
Escape - give up, make no response.
High levels of stress and anxiety can increase these effects. The situation had progressively deteriorated from the time of high speed initial approach, and the subsequent actions not achieving the desired results. It is also probable that the captain’s level of stress and anxiety had progressively increased as the initial approach, and then the orbit, did not go as he had intended.
As said before, there are no similarities between the accidents, this post only should point to the fact, that pilots are no supermans and that human errors are also mostly systemic errors.