Investigating SPADs - Four things to remember
This article provides guidance on four steps which aim to focus members’ efforts:
- Follow your company’s investigation procedure (Safety Management System)
- Identify the full range of underlying causes, not just immediate causes
- Classify causes using RSSB’s 10 Incident Factor framework
- Prepare effective recommendations
See below for the detail.
1. Follow your company’s investigation procedure
Make sure you follow your company’s investigation procedure - it will help you cover the basics. It’s also worth familiarising yourself with the Rail Industry Standard for Accident and Incident Investigation, RIS-3119-TOM. It provides you with detailed guidance on conducting an investigation, interviewing skills, identification of underlying causes and so on. Appendix P also provides some SPAD specific considerations, for example:
- Types of SPAD
- Evidence to gather following a SPAD
- RSSB’s SPAD Risk Ranking Tool
- Signal Sighting Committees.
2. Identify the full range of underlying causes, not just immediate causes
Analysis of SPAD investigations shows that they are good at describing the immediate events and issues that led to the SPAD. But getting to the bottom of a SPAD, to the underlying causes and making robust recommendations, is something where further guidance is needed. These two fictitious summary SPAD investigations show how to get to the required depth.
For each factor that directly led to the SPAD, ask yourself ‘why did that happen?’, and keep asking ‘why?’ until you get to underlying, and ultimately, root causes. Getting to the root cause is important because this will enable you to make useful recommendations that can help improve safety in the future. For example:
- Why did the driver pass the signal at danger? Because they read across to a signal on the adjacent line.
- Why did the driver read across to the wrong signal? Because they were not absolutely familiar with the route having not driven over it for several months.
- Why had the driver not driven over the route for several months? Last time they were due to drive over the route their train was cancelled. This was not fed back to the driver’s local manager who was dealing with a higher workload than usual because they were covering both their normal role and a vacant post.
- Why was the driver’s local manager covering for a vacant post? And so on...
In the example above, rather than just focusing on the driver’s error which was the immediate cause of the SPAD, asking ‘why?’ several times means we have been able to establish underlying and root causes. Recommendations may centre around improving reporting channels to ensure local managers are made aware of times when drivers have missed rostered trips over infrequently-used routes, and about expediting manager recruitment so that their workload is not excessive.
It’s important to think about issues beyond the driver and their immediate actions. For example:
- Were there any equipment issues?
- Were there any unusual signalling sequences or layout?
- Were organisational issues involved, such as unusual shift patterns, or route knowledge problem?
Use the 10-incident factor framework, described below, as a checklist to help make sure you have covered all the bases.
3. Classify causes using RSSB’s 10 Incident Factor framework
Classifying causes will help identify the root and underlying causes of SPADs and helps to prepare more effective recommendations. The 10-incident factor framework (see diagram below), separates human performance issues from 10 factors that can contribute to these. For example, a driver acknowledges an Automatic Warning System (AWS) warning, but rather than slowing down, the driver incorrectly takes power, which is a human performance issue. This could have been due to fatigue, workload or environmental issues (three of the 10 incident factors) and each would require different recommendations. Further details on the framework, including sub-categories can be found in the Rail Industry Standard for Accident and Incident Investigation (section H.7, page 58). RSSB has prepared guidance on how to classify incident factors using the 10-incident factor framework.
4. Prepare effective recommendations
These will aim to prevent similar SPAD events occurring by addressing the causes of the incident. A good question to ask yourself when writing recommendations is: ‘will this change the behaviour of one individual, or will it affect the actions of many?’ To illustrate, improved signal sighting will help design-out a problem for drivers in the years to come, whereas issuing a SPAD alert may raise awareness, but effects will be short-lived. Reviewing incidents at the location and with the driver are separate activities.
When human performance issues need to be addressed, it’s important that recommendations are linked to the types of causes identified. For example, following a depot SPAD, the investigation identified the driver was not familiar with the depot layout signal positions which led to the incident. Recommendations for this scenario should focus on ensuring all drivers understand the signalling and depot layout. It might also be tempting to recommend the driver adopts techniques to maintain ‘situational awareness’, but if a ‘loss of attention’ was not identified, it will bring limited benefit. Appendix K of the Rail Industry Standard for Accident and Incident Investigation has guidance on how to link causes with recommendations and has useful advice on writing recommendations. Getting this right helps underline the principle of a fair culture, as well as producing effective recommendations.
On the other hand, findings from an investigation may suggest a need to develop a driver’s competence. For such cases, RSSB has prepared guidance on competence development plans. Some examples where competence development might be useful include:
- Route or traction knowledge issues have been identified
- Issues with spoken safety-critical communications and not reaching a clear understanding with others
- Driver not prioritising shunting moves effectively, suggesting work on ‘task organisation and prioritisation’ is necessary
When deciding to adopt a competence development plan, think more broadly than the individual concerned. Is it likely that others doing the same job might also respond in a similar way / take a similar course of action? Could the issue reflect a wider problem, requiring a broader intervention?