State misinterpretation has been identified as a causal factor in several accidents where humans were operating complex systems in dynamic domains. The concept of state misinterpretation, although undefined, is characterised by its unobservability, and its relative infrequency. These features make gathering data about state misinterpretation difficult. It was therefore decided to use archive incident report data; the NASA Aviation Safety Reporting System database was used. A definition of state misinterpretation was formulated and translated into database queries to retrieve relevant incident reports for a homogeneous set of expert pilots over a fixed time period. These reports were encoded using the Cognitive Reliability and Error Analysis Method suitably adapted to analysis of aviation incidents. Each report was categorised by the type of state misinterpretation, and a taxonomy of these types was developed. Those types which occurred more than 20 times were analysed at three levels of abstraction. First, a concordance of individual actions showed that communication failures, missed observations and distractions were the most common causal factors. Second, the sequences of possible causal actions showed that some sequences are common across different types of error and state misinterpretation. Third, the causal trees for each state misinterpretation type were quantitatively compared. The lack of measured similarity between the trees suggests that the types in the taxonomy are distinct. Most of the analysed incidents were preventable by better management of flight crew actions. Two particular sequences of actions dominated the results. The first is where the flight crew missed an observation when they were distracted by a competing task. The second is where a communication failure between the flight crew and air traffic control occurred. Some suggestions are offered about how flight crews can better manage their actions to prevent the occurrence of some types of state misinterpretation, thereby reducing incident numbers.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:364665 |
| Date | January 2001 |
| Creators | Baxter, Gordon |
| Publisher | University of Nottingham |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://eprints.nottingham.ac.uk/13645/ |
Page generated in 0.0018 seconds