Health and safety data: quality information drives quality decisions

This month we're joined by Marcus Dacre, RSSB’s Head of Risk and Safety Intelligence.

He talks with us about the importance of having good quality health and safety data in the rail industry. Marcus also covers the work that got us to this point, what the future holds, and the improvements we're making to SMIS to aid better data capture. He finishes up with a discussion of some standout stats from our 2024 Annual Health and Safety Report.

Find out more at rssb.co.uk/risk-and-safety-intelligence

Transcript

Host [00.15]: It’s 2024. And in more ways than one, the future that many fantasised about back in the 50s and 60s—one built on automation and intelligence, and on endless possibilities—has become our present. The pace and scale of many recent technological advancements have transformed our industry, and we now have more data than ever before. But how do we use that data to maintain our strong health and safety record? 

I’m joined today by Marcus Dacre, RSSB’s Head of Risk and Safety Intelligence, to talk about the importance of good quality health and safety data in the rail industry.

Marcus, hi, and thanks so much for joining us today. 

Marcus Dacre [00.59]: Hi, Sarah. Thanks for having me. 

Host [01.01]: Let’s get straight into it. How do we use data in rail? 

Marcus [01.06]: So the railway’s always generated lots of data and used that data to improve. So, my particular focus is health and safety. And you can find facts and figures about rail accidents and casualties in HMRI reports going right back to early Victorian times. Of course, over time, we’ve got much better at using data. And in recent decades, GB rail has developed and benefited from an evidence- and risk-based approach to safety management. So, we collect good safety data, we share it, we use it to inform safety-related decisions, and we use it to give us confidence that those decisions are protecting our workforce and the people that use the railway that we’re meeting our legal requirements for safety and we’re also meeting wider business objectives.

Of course, we’re living now in the information age, where we’ve been seeing a massive growth both in terms of the volume of data we’re producing and the value we’re getting from it. And that applies to health and safety as much as any other topic. 

Host [02.01]: And why is it so vital to have good quality health and safety data in our industry?

Marcus [02.05]: I suppose the simple answer is that if we want to make high quality decisions, we need high quality information. And we use that information to identify new or changing risk. We use it to help focus our effort and resource on the right areas. We use it to decide whether we should or shouldn’t invest in a particular risk control. And if there are gaps in our understanding of risk caused by missing, incomplete, or incorrect information, then we’re unlikely to be managing health and safety as effectively as we should be. 

I suppose one point of clarity is that I think when we’re talking about data today, we’re tending to think about, I suppose, hard facts and figures. I think it’s important to recognise that information comes in lots of forms, so from the conversation between a manager and a member of frontline staff, through to an accident investigation report, through to the feed from a platform CCTV camera. I think all of those are potentially important pieces of the information jigsaw. 

Host [02.59]: Thanks, Marcus. Could you now tell us a little about the current health and safety data landscape?

Marcus [03.06]: So, generally, the companies that work in GB rail have got really mature and effective safety event reporting systems. And they also share a lot of that data through a system called SMIS, the Safety Management Intelligence System. And this allows organisations to learn from each other’s experience and pool safety data across the 30-odd operators in the mainline railway. So that provides a much larger and richer source of information that everybody can use.

In the safety space, people are also getting better at recording not just accidents, which is where things have gone wrong, but also close calls. So that’s unsafe acts or conditions that didn’t lead to anyone getting hurt, but perhaps could have done in different circumstances. In health and well-being, we’re less mature in terms of data collection and management, but we’re currently building the foundation for that evidence- and risk-based approach that I spoke about earlier.

In RSSB, we recently completed a project where we worked with a number of companies to develop health and well-being dashboards. We’re now looking to build on the success of that and really make that part of business as usual across the railway.

Host [04.10]: And what about the work we’ve done, and the work the industry has done, to get to this point? Can you touch on that? 

Marcus [04.17]: Yeah, perhaps I’ll focus on SMIS in this answer, which my team manages on behalf of the industry. We can trace the origins of that system back to the Clapham Junction accident in the late 1980s. And the first SMIS system with that name came into being around a decade later. And we’ve developed a long way since then, and the system has evolved alongside the culture of reporting and open sharing of information between operators that’s fundamental to it.

So, at RSSB, we provide guidance and support to users of the system. We work closely with our member companies who use the system, and they advise us on both the day-to-day running of it and its long -term development.

We’ve got a really extensive data improvement programme to support the data quality, which we touched on earlier. So SMIS users have access to a live report that flags the potential data issues. And we carry out data quality health check meetings twice a year with most SMIS user organisations. And these take stock of how they’re doing and where we can support. And at the end of the year, we provide each organisation with a data quality scorecard and an assessment of their data quality in the system. And this focus on data quality has really helped to improve standards over recent years.

The current specific area of focus is on accident causes, and I think it’s probably traditionally fair to say that SMIS has been better at recording what happened and where and when it happened than necessarily why it happened. So, we’ve developed an improved causes module, and our human factors team is currently working with operators to improve the information that’s being pulled from their investigations and shared through SMIS. 

Host [05.53]: Thanks, Marcus. I think it’s definitely an area where there are always improvements to be made. So, how will we continue to strengthen this type of data and make the most of its insights?

Marcus [06.04]: Yeah, so again, I’ll stick with SMIS probably for this answer. We’ve got a major programme of work under way to make the system quicker and easier to use and to help people get better intelligence from it. 

We recently finished a piece of work to simplify the entire SMIS data model. And that’s already generated significant time savings for our members. And it’s also an enabler for the major project we’ve now got under way to build a new Safe Insights platform that’s going to replace the current SMIS solution.

So, we’ll have the new system ready to pilot in autumn. We’ll complete the transition to the new system by the end of the fiscal year. In that system, we’ll be using artificial intelligence to speed up the input process, to improve the quality of the data, and to generate insights from that data. And we’ll also be providing functionality that allows automatic transfer of data between company systems and the central SMIS system.

That’s probably the input side. On the output side, we’re doing a lot of work too to maximise the value that people can get from the information in SMIS. So, we’re improving how operators can access that data, for example, to create their own reports and dashboards. In terms of the outputs that RSSB produces, we’re moving away from the traditional static reports that we updated maybe periodically to more dynamic data hubs that work from the live SMIS system. And these allow decision makers to slice and dice the data as they want and drill down into it so they can really better understand the trends and what’s driving them and so how to address them. And a number of those data hubs are already available on the RSSB website.

I probably should also mention the risk modelling work that we do. So, high consequence collisions and derailments are very rare on today’s railway, we have a good safety record, but it’s vital that we continue to understand and manage the risks from those very low frequency but potentially catastrophic accidents.

So, we’ve got a mature and well-used Safety Risk Model, and that uses data from SMIS and from other sources, alongside sophisticated modelling techniques and expert input, to structure and quantify those risks. And a real strength of our Safety Risk Model is that it’s grounded in the reality of what’s happened, but it’s not constrained by it. So, it covers the risk from accidents that have never occurred on the GB mainline, alongside those like the trips and falls in stations that are still occurring every day.

Host [08.24]: And what does the future hold for real health and safety data, more broadly? 

Marcus [08.29]: Yeah, it’s a really, really big question. And there’s lots of excitement at the moment around big data, machine learning, artificial intelligence. And that excitement extends to health and safety management too. And it’s not just about tomorrow. We at RSSB and many others across the industry are already exploiting new data sources, new tools and techniques, and delivering genuinely new insights that are helping improve how we manage health and safety risk. 

I suppose what I would emphasise though is that having really good safety event data, of the type we collect in SMIS, remains important and is arguably more important than ever, because now we can get much more value from that information by combining it with other data on railway operations and engineering, railway assets and their condition, the environment that we’re operating in. 

And there are probably four broad areas where we’re driving improvement, seeking to provide genuinely new insights into risk and how we can manage it. Bringing data-led decision-making closer to real time, so we could be more proactive in how we’re managing health and safety and intervene before something goes wrong, not just reacting when it does. I think we can provide much more granular information now than we could do in the past. So, the data that informs a specific decision is much more specific to the particular location and the issue that’s of interest. We’re less reliant on those broad-brush assumptions that perhaps we had to make previously. And finally, I think something about supporting more holistic decisions. So, a decision that affects health and safety isn’t made in a vacuum; the railway is a complex system, and that decision may well have a knock-on impact on other aspects of health and safety. It’s also very likely to affect other dimensions of risk, like performance and cost. So, it’s really sort of being able to support much more holistic decision-making. 

Host [10.15]: Thanks, Marcus. We’ve also just published our Annual Health and Safety Report, which unpacks the risks across the network and the work that’s under way to avert them. Can you talk to us a little about that? What areas does it look at, and what value does it bring? 

Marcus [10.32]: Yeah, so I’ve been involved in the Annual Health and Safety Report and its predecessors for quite a long time. It’s our annual stock-take really of health and safety performance on the mainline railway. We structure it around the five key risk areas in the industry’s Rail Health and Safety Strategy. So, those are health and well-being, operations, occupational health and safety, asset management, and public behaviour. There’s an awful lot in there, and I’d encourage people to go and take a look at the document and all the supporting information that we publish alongside it. 

In terms of what it’s for, it helps the industry groups who oversee the delivery of the strategy in each area to make sure they’re focused on the right thing and to track the impact of the initiatives they’re leading as well as other changes. I think individual operators can use it to benchmark their performance against what’s happening in the wider industry, and they can use it to identify relevant health and safety activities and resources. The regulator uses the information to inform its priorities. 

And finally, I think transparency is really important in health and safety. So, we publish a report so that passengers, freight customers, tax payers, and the groups that represent them can understand current levels of health and safety risk and what we’re doing to manage it.

Host [11.45]: And are there any standout stats in this year’s Annual Health and Safety Report? 

Marcus [11.50]: When we’re pulling the report together, and this is an activity we carry on through the year, we’re really acutely aware that every injury statistic relates to a real person with a life, hopes of a future, friends, and families. So, I think by far the best statistics are the zeroes. And one that stands out from last year’s report is that there were zero workforce fatalities, which is great news. Sadly, it’s not the norm. We already know that tragically we won’t be able to say the same for next year’s report. But lots of people from across the industry are working to reduce occupational health and safety risk and bring us closer to meeting that zero harm aspiration.

The report has also got some interesting analysis that highlights the challenges that we’re facing as a result of changes in risk that come from changes in society and public behaviour and also a changing climate and the prevalence of extreme weather events.

I always find the Safety in Context report that we publish alongside the main report really interesting. Like I said, lots of the analysis necessarily focuses on things that have gone wrong. I think the Safety in Context section allows us to step back a bit and understand how much things have got better over recent years and also how the railway compares to other international railways and other modes of transport. It shows that, really, GB rail has got a really good safety record, albeit one we’ve always got to strive to improve on. 

Host [13.13]: Thanks, Marcus. And finally, do you have any advice for listeners whose role might include reporting health and safety data?

Marcus [13.22]: I suppose, first of all, for those listeners who are listening, I’d like to say thank you. That aspect of your role is really important. Hopefully, I’ve covered it already, but good reporting is really vital to good health and safety management. And the value of that reporting extends beyond getting the immediate issue fixed. So, by recording complete and accurate information about what happened and why, other people can learn from that and perhaps prevent similar issues from arising in future. And also, senior leaders and funders can make strategic decisions that are informed by a good objective understanding about what’s happening on the ground. 

I think people that have been around for a long time in GB rail, we sometimes take SMIS for granted, but when you talk to people in other sectors, from other railways, they’re really blown away by the information that’s in the system and the value we get from it and the culture of sharing that supports it.

Probably that’s not advice. I suppose, in terms of advice, we’re here to help. So, if you’ve got a question or you’ve got an idea about how we can improve, then please do get in touch with us. 

Host [14.22]: Marcus, thank you so much for joining us today. And thank you for listening. If you want to learn more about health and safety data and the Annual Health and Safety Report, please visit our website.

We look forward to you joining us for the next episode, and in the meantime, safe travels.