Asset integrity – it’s not all in the data
I was at a seminar once, listening to some very big and important people. One of them said safety was ‘all in the data’. He was right…and yet he was also wrong. The trouble with data is that it’s only as good as what gets captured, what gets analysed and what gets digested by those who make the decisions.
In the late 1980s, British Rail was focussing on signals passed at danger (SPADs), as that’s what was worrying it most. The 1988 collision at Clapham - which killed 35 people and injured many more - resulted from a signal showing green when it should have been red to protect a train ahead. There’d been several similar ‘wrongside’ failures (as we call them) before this one. The data was there, but it was dispersed. No one gathered it together, because the organisation was looking elsewhere. If they had gathered it, a trend would have been revealed, which could have led to an alert, which could have saved all those lives.
Come 2000, the problem had seemingly shifted to the integrity of the track itself. At Hatfield on 17 October, an express train heading for Leeds derailed on a curve travelling at 115mph, killing 4 people and injuring 70 more. The track had shattered. The immediate response was to impose stringent speed restrictions across the network while the state of this assets was analysed. The resulting inquiries pointed to a number of issues, including a lack of asset knowledge, lack of appropriate training and a culture that put performance over safety. There was also a data issue, it being revealed that there was no system within Railtrack that allowed the company to view the total number and types of rail defects affecting the infrastructure at any one time.
One of our bigger threats now sits in the digital world, as evidenced by the loss of temporary speed restrictions on the Cambrian in 2017. This highlighted problems not only with the signalling software in use on the line, but also our incomplete understanding of the problems that can go with it. The tragic accident at Carmont (2020) and the derailment at Llangennech (2020), remind us that digital is only one of the challenges asset managers face, the former resulting from a landslip after heavy rain, the latter involving the condition of a tanker wagon. All three incidents show the need to have ‘eyes everywhere’, and the need to bring these ‘eyes’ together into a single all-encompassing viewpoint.
But what about the interface between the digital world and the traditional ‘nuts and bolts’? Train door mechanisms and vehicle loading monitors sit in this category—and it’s a category that will only get bigger and more complex.
So, can we really be sure that everything is in the data? Is everything finding its way to SMIS and NIR Online, for example? Formal investigations can shed more light, of course—they’re thorough, scientific and objective; they give us the detail we need. Yet they take time to complete.
We do have Close Call information, but there’s a huge amount of it, making it fairly difficult to mine. What, about the daily reports - what indeed about conversations? In any depot, station, or corridor, our colleagues will voice their concerns as they take a break or get a coffee. As the end users, they’re the closest to it every day - at linespeed. We need to talk to these people more often - and we need to listen. This, with the qualitative information and the stats we already have, and the investigation reports we will eventually get - will help us construct the grand narrative we desperately need to build a better, safer railway.
Dr Greg Morse is RSSB’s Operational Feedback Lead and the Editor of the Leading Health and Safety on Britain’s Railway strategy. Greg uses his storytelling skills to capture learning from operational experience and provide the railway with the tools to retell those stories among staff at all grades. He is a Member of the Institution of Railway Operators and has written many books and articles on railway history in his capacity as a professional author.
If you want to know more about LHSBR – please get in touch.