Written by Dr Anjum Naweed
With a combined resource total of more than $100 million to be utilised over seven years, the Cooperative Research Centre (CRC) for Rail Innovation is a significant asset to the Australian Rail Industry.
The CRC is a unique and productive endeavour linking the substantial skills and resources of the rail industry with the research and development expertise of seven of Australia’s leading universities.
Collaborative research enables a small contribution from individual rail companies to be leveraged in order to solve serious problems common to several Participants.
This article is based on the work being conducted in the Safety and Security stream of the CRC for Rail Innovation’s research projects.
As the title suggetss, the projects in this stream are concerned with all things safety, and range from the tehnical, such as the next-genereation of fatigue risk modelling, right the way through occupational health and safety concerns in cultural management.
This particular article describes research that falls somewhere in between; it traverses the pin-point of the sharp-end and concerns itself with the human at the head of the train.
It may not seem it but train driving is an astonishingly complex task – guiding tonnes at breakneck speeds, while ensuring that you don’t go into unauthorised track- that’s relatively easy but when you add the physics of driving on metal, you have a task that the driver may only control under a continuous stream of two time spheres: an awareness of where they are and where they need to be.
Of course, the railway’s temperament is affected by visibility and adhesion too – a wall of fog, a smattering of dew, some poorly felled leaves, or an ill-fated millipede can completely change the game.
Then there’s the type of operation to contend with – you want to avoid endangering the passengers of course, and getting into the station too late, or eating into your dwell time, and bunching the train, or dumping the air, derailing a wagon, having train control ring you up every moment – the list goes on.
Then there’s the sheer openness of the railway – what obstacle or impediment is waiting round the next curve, at the bottom of the hill, in the upcoming tunnel, at the next level crossing?
And let’s not forget the myriad of human factors – Issues born from having to maintain speed under such specific circumstances: fatigue, shift work, sustained attention, vigilance, reaction time, and so on – all part of the railway’s endearing personality.
Ah yes, train driving is a complex task, and train drivers must maintain an accurate awareness of where they are and what’s coming up, so that the speed and signalling-based constraints of the railway fall inside the train’s tractive and braking limitations. Some train drivers live in the proverbial past as well as the future.
Heavy haul drivers, for example, need to know the precise location of the tail-end which can be located anywhere up to 2 miles behind them.
All this necessitates route knowledge, a detailed body of information that train drivers are expected to cultivate and a singular peculiarity of the railway system that emerges from harbouring so many rich and diverse driving factors.
In practice, this involves an extremely detailed and dynamic understanding of one’s route, including the trackside furniture, prospective landmarks and their fixed and/or relative locations to one another, in order that braking distances may be effectively predicted and throttle action may be properly regulated.
There is a good deal of research that has explored the route knowledge process and mapped out its contributing factors but very little work has sought to understand how the route is actually encoded in the driver’s mind. How do train drivers actually picture their routes? How does this change between operation types and does it vary as a function of expertise?
These are basic questions but extremely important ones, and identifying how the route is mentally structured would allow us to better understand the nature of the task, and by extension, better inform the design of training programs.
A key part of the CRC’s Safety and Security research stream are the capturing driving strategies and route knowledge acquisition projects. We’re leading these in conjunction with the CRC’s rail simulation laboratory and have designed them to explore methods for enhancing driver learning and advancing competencies.
These projects aim to derive a better understanding of route knowledge, in particular the way that routes are encoded, so that this information may be captured and delivered effectively through simulation.
However, much of this knowledge is intuitive and not always available to conscious introspection, so how do we collect it? How do we reach into the thoughts of such a diverse population, and snatch such complex information right out from under their mind’s eye?
As researchers, we need to start thinking outside the box a little, so we’ve developed a novel framework purposefully designed to elicit some of that deeply buried intuitive knowledge.
The approach is qualitative, which means we’ve favoured some of the more traditional methods, such as interviewing and observational techniques.
We’ve spoken to train drivers with varying levels of experience, in groups and on a one-to-one basis, but in a new twist, incorporated some generative tools traditionally used in participatory design.
During our interviews, we’ve asked drivers to invent difficult routes – to actually draw them – and then to use their drawings to describe scenarios that could challenge even the most seasoned train driver. Of course, we’ve spoken with them about real routes too and explored any formative experiences, so that we can chart how their skills have evolved.
We’ve also carried out observations in the train and asked drivers to think-aloud and provide a commentary for all their actions.
This is a process that represents the contents of the driver’s working memory – obviously, there are a lot of individual differences in its articulation, but it gives us some idea about what cycles through the driver’s mind.
Clearly, extended periods of no commentary tell us just as much as the talking, and provide us with information about periods of high workload or automatic driving behaviours.
An example might be: “Departing station at notch 2, notching up to reach track speed, watching absolute signal, focussing on absolute, checking speed, powering back into coast, checking signals are correctly set, around the corner is a level crossing, pedestrian maze and permissive signal, checking signal, around the next curve is pedestrian crossing and large station, checking speed, applying some brake, level crossing is 300m away, just over crossing and around the corner is absolute signal blind to you, multiple lines diverging because of…”
We’ve also tried to take each train driver systematically through these phases; having them explore the topic as a group, then on a one-to-one basis, and then in a cab, not only increases our understanding of their task but also gives them an opportunity to find their words, as it were, and really contextualises the whole experience.
It’s a fundamental part of the process – drivers have taken us over routes that they’ve talked about and illustrated in the interviews, allowing us to explore ad compare the symmetry of their psychological railway with the actual railway.
So far, we’ve conducted our research with passenger and heavy haul train drivers in New South Wales, South Australia, Victoria, Western Australia, and Queensland. We know that this methodology is working because at the end of a day’s data collection, novice and expert drivers alike are exclaiming that they’ve “learned things” about what they do, and “discovered insights” into their task.
We’ve only started eyeballing the data at the moment but it’s saying that train drivers are mentally organising their routes in a manner far beyond what we originally conceived. Some of these include very sophisticated 3D coordinate representation systems that dynamically connect the railway’s surface shape with its gradient profile.
It’s also really interesting to note that perceptions of route difficulty appear to interact with perceptions of enjoyment, thus, difficult routes appear to be more enjoyable to drive.
The next step is to finish up with the data collection and then start on its analysis with an aim to use this information to optimise the driver training process.
To clarify an earlier point, one of the best ways to achieve this is though the effective use of simulation. To that end we’ve also visited a number of rail simulators in the industry and discovered that, whilst there is much variation in the way they are being applied, most are highly underutilised, creating a lot of untapped potential in the use of these tools.
Once we’ve got a good idea of the way that route knowledge is internally represented, and how this interacts with navigational strategies, we will be able to translate this to the simulator and start investigating scenarios that enhance learning and optimise competencies.
It’s really exciting work that will help improve the safety of the rail system and protect the welfare of the train driver, and further down the line, bring us one step closer to harmonising the Australian rail industry.
For more information about the research of the CRC for Rail Innovation please visit their website at www.railcrc.net.au
Dr Anjum Naweed is deputy program leader for ops and safety, CRC for Rail Innovation and postdoctoral research fellow at CQUniversity Australia.
