Systematic Human Error Reduction and Prediction Approach
Systematic Human Error Reduction and Prediction Approach (SHERPA) is a human error identification approach, originally designed to assist those working in the process industries (e.g. conventional and nuclear power generation, petrochemical processing, oil and gas extraction and power distribution; Embrey 1986) to better understand error potential. The domains of application have broadened in recent years, to include areas such as aviation (Harris et al. 2005) and health care (Lane et al. 2006, Phipps et al. 2008).
SHERPA is based on normative models of task performance, such as HTA. Where HTA has been completed at the systems level as discussed previously (i.e. where ‘operators’ have been defined as including both human and technological actors), SHERPA analysis enables an understanding of potential failure across all aspects of the system. SHERPA uses a taxonomy to classify different types of potential errors. The taxonomy is based on the following five task types:
- 1. Action (e.g. pressing a button, engaging a piece of equipment, opening a door)
- 2. Retrieval (e.g. retrieving information from a display or manual)
- 3. Checking (e.g. conducting a check for signage)
- 4. Selection (e.g. choosing one alternative over another)
- 5. Information communication (e.g. exchanging information through verbal or non-verbal means)
The outcome of a SHERPA analysis is a set of credible errors that can be prioritised based on probability and criticality, with concomitant error reduction strategies or interventions.
BOX 2.10 HUMAN ERROR IDENTIFICATION FOR RAIL LEVEL CROSSINGS
In this research programme, we used SHERPA initially to understand the potential errors associated with the current system (see Chapter 5) and then subsequently within a design evaluation and refinement process to assess how new designs would address the existing errors or introduce new errors (see Chapter 8).