Predicting pilot error: testing a new methodology and a multi-methods and analysts approach.

Area of Science:

• Aviation Human Factors
• Pilot Error Analysis
• System Safety Engineering

Background:

• Pilot error is a significant factor in aviation accidents.
• Accurate prediction of design-induced pilot error is crucial for improving flight safety.
• Existing Human Error Identification (HEI) methods vary in effectiveness.

Purpose of the Study:

• To validate the Human Error Template (HET) methodology against contemporary HEI approaches.
• To compare the predictive performance of HET for pilot errors in approach and landing tasks.
• To evaluate the effectiveness of a multiple analysts and methods approach for enhancing error prediction sensitivity.

Main Methods:

• The study compared the Human Error Template (HET) with three other Human Error Identification (HEI) techniques.
• Pilot errors for an approach and landing task were predicted using each methodology.
• A multiple analysts and methods approach, pooling predictions from various techniques, was also assessed.

Main Results:

• The Human Error Template (HET) demonstrated the most accurate pilot error predictions when used in isolation.
• The multiple analysts and methods approach showed higher error prediction sensitivity than individual methods, except for HET alone.
• HET proved superior to other individual HEI methods in predicting design-induced pilot errors.

Conclusions:

• The Human Error Template (HET) is a highly effective tool for predicting design-induced pilot error.
• Employing a combination of different HEI approaches and multiple analysts improves error prediction sensitivity.
• A toolkit approach integrating various HEI methods and analysts is recommended for robust error prediction.