’Minimising the risk of a SPAD’, originally published in Track & Signal October-December 2015. Written by Anjum Naweed, Principal Research Fellow at the Australasian Centre for Rail Innovation (ACRI).
For all the technical complexity underpinning them, railway tracks possess a rather simple and elegant truth: rolling stock can move only in the direction they provide. And so it is that signals have come to adorn the trackside of conventional rail networks, showing the driver where they rare authorised to go (and in some cases, how fast), creating a unique multi-aspect dynamic that few collision avoidance domains can boast. The dark side of the equation of course is the SPAD (Signal Passed At Danger).
To the general public, a SPAD is a newspaper headline that strikes terror in the imagination and, for some, conjures concerns for railway safety. To the knowing passenger, a SPAD is an unlikely statistic that surely won’t happen to them on their commute; to the general operations manager, a SPAD is long hours spent deliberating over the rail network, entertaining hope but bracing for inevitability; and to the train driver, a SPAD is an emotional response beyond words. All this in a four-letter acronym that means “a signal has been passed at danger”.
A SPAD is the single biggest failure mode possible in the railway and straddles the fine line between accident and disaster. The rail industry is emblazoned with it, across the driver-cab dashboard or on the notice board in the drivers’ mess hall, such that it seems to have taken a life of its own. Technically, SPAD describes the event of a train or locomotive encroaching into a section of unauthorised track. In practice, this can be with or without the driver’s knowledge and through an error of omission or commission, giving rise to a plethora of human factors issues.
Say that a train driver is distracted by a pedestrian at a crossing, delays braking and has a SPAD; in this example, they apply the brake later than intended and in doing so perform an error of omission. On the other hand, say a loco driver goes unwittingly through a red signal when departing a yard; that would be an error of commission. Some SPADs are more ambiguous: say, a driver is approaching a stop signal, brakes by the book but then suddenly encounters a ’plague’ of fearless millipedes swarming over the railhead. The driver applies the brake as best they can but ultimately doesn’t regain control and slides straight past the stop signal. What kind of error is that? Technical perhaps? Tricky one. Of course, the most serious kinds are those in which the train sails past a signal and the driver remains none the wiser that they just went through was a stop aspect. This is a simplistic way of looking at what is quite a complex failure mode, but what is important is realising that for one reason or another the signal and the driver have become disconnected.
Nowadays, safety technology has been woven into the fabric of the railway to help mitigate SPAD-risk. Most advanced conventional rail networks use ATP- or TPSW-type systems which work to stop a train if it passes a danger signal, but by and large these methods serve only to dampen the effects of a SPAD; they arrest speed once the SPAD is performed so do not address the cause itself. For this reason rail networks also use systems to mitigate risk at the driver-processing level. Some of these are set/reset devices like the old Automatic Warning System that forces the driver to acknowledge cautionary zones while others supervise train movement and prompt remedial braking for excessive speeds (for example, ERTMS Level 1).
All of these countermeasures attempt to address the latent human factors issues born from operation in a heightened and sustained state of attention for so long. But what the industry must understand is that to the train driver, the railway is a little bit “Jekyll and Hyde”: it can manifest in the form of an ostensibly harmless and monotonous landscape with few signals and little change in information or, conversely, as a dynamic world with many events and monstrous infrastructure densities. Driving in both of these states poses challenges for driver situation awareness, distraction and mental fatigue, and the problem is that while forces-response-type countermeasures may work to mitigate an element of risk, they can also exploit other human factors issues, such as automaticity and habituation.
The topic of SPADs has been the subject of much research (and contention) but findings have always reached an impasse. We know to a certain extent that any disconnect between the driver and the signal may arise from signal sighting issues, poor driving conditions and inattention, but less clear are the SPADs that result from signals being misread or misjudged and from the driver getting distracted in their own ’head-space’. We know that a whole host of human factors behavioural issues play a part but what we don’t know much about is how drivers cope and condition themselves to mitigate the problem at an individual level.
Last year I completed a project funded by the Australian Cooperative Research Centre (CRC) for Rail Innovation that looked beyond the confines of technology to the strategies that drivers adopt to manage SPADs. The project is now in various stages of uptake but it is a good time to share with the larger rail community some of the insights we have gained. Most train operators (passenger, freight and heavy-haul) in Australia and New Zealand were involved in some capacity. The aim of research was to determine the underlying behaviour-shaping factors that affect SPAD-risk by examining how train drivers engaged with risk under different conditions. A mixed-methods research design was used to examine difference SPAD-scenarios, and a comprehensive analysis of these scenarios has been used to develop a model of risk for subsequent quantification of error-producing conditions. I will come back to the model in the next issue of Track & Signal, but here I will touch on some qualitative data that readers may find interesting.
When collecting data, this study asked drivers to “describe their relationship with the stop signal” - that is, articulate how they saw the signal. This may seem an odd question to ask, particularly from the perspective of the hardened rail manager, but put yourself in the train driver seat for a moment. You’re driving over the railway, day in and day out, on your own for the most part, watching it snake endlessly before you and vanish on the horizon line. The sleepers blur into one another, though you can count each one if you drive slowly enough. The one thing you engage with all day every day in its droves - hundreds if not thousands - is signals. Each signal has an aspect telling you if it is safe to move, and each signal is used to inform an ever-evolving awareness of the world, whether to sustain your knowledge of the route or inform throttle and braking actions, or both. Signals have an important part to play in your life so it stands to reason that you will also have a unique relationship with them. When it comes to learning about SPADs, what can your relationship with signals tell us about the way you view risk?
The question that sought to ask train drivers about their relationship with the signals was completely no-holds-barred so drivers answered it as they saw fit. One group of answers were what you would expect - testimony that signals were, and are, the “top priority” for driving and represent that “ultimate collision avoidance system”. But drivers also described signals with different levels of personal intimacy. For instance, one kind of relationship was all about a sense of regard and high-esteem that drivers had for signals, with choice words like “respect” and statements like “it’s the most respected thing out there”.
Another kind of relationship projected the effect of adhering to signals on the driver, or on their passengers, or on the rail network as a whole: “my life-line”, they said, or “my passenger’s safety” in recognition of the role that obeying the signal played in the grand scheme of things. Some drivers described the impact of their relationship with the signal in the broader context of their lives, and “my livelihood” and “my bread n’ butter” were often repeated.
Interestingly signals were also personified. This is not altogether surprising, given how isolating driving trains can be (in spite of many bodies walking into and out of carriages all day). This personification occurred in the form of a direct driver-signal dynamic where the signal was described as “my colleague” and “my best friend”.
Another type of relationship extended the reach of the signal to that of a frontier and “a boundary” or “a brick wall”. Lastly, the relationship went beyond personification into a space altogether more divine and the signal was described as “my religion” and “God”.
Myriad answers all reflected a unique property of the signal in the architecture of the railway: that beyond the lights, LEDs and colour, the signal is associated with many psychological constructs. On this level, the answers revealed that signals were something to be relied on, befriended, respected, trusted, feared and obeyed. The signal was all of these, but behind it all lurked the SPAD and the notion that one day a driver may accidentally make the wrong prediction or, somehow, violate the tenets of their relationship.
The intensity of the driver-signal relationship is such that violating its terms and having a SPAD created symptoms on an acute stress reaction (release of noradrenaline, increased hart rate, constricted blood vessels, change in blood pressure).
This was accompanied by extreme emotional irregularity (disbelief, fear, panic anxiety).
Given this, what would it mean for those who describe a signal as “my religion”? Does it amount to a sin? What happens after a SPAD? Does their relationship with a signal change from it being a friend to a brick wall? For their top priority to their livelihood? These are all important questions from a phenomenological standpoint.
So you see, it is actually quite important from a human behaviour perspective to seek data for these sorts of issues before it is possible to understand underlying causation, risk engagement and the nature of SPAD-mitigation. The scenarios produced by the drivers were used to identify the key contribution risk factors that were considered to significantly increase SPAD-likelihood.
(Source: Australasian Centre for Rail Innovation - ACRI)
