Last Updated on August 23, 2026
Dr Ben Guy, Urban CGI — Urban Consulting Group International
There is a number inside most rail operators that nobody owns. It does not appear in a business case, it is never presented to a board as a single figure, and it is large enough that if it were a capital item it would have its own governance. It is the cost of getting train crew route competent.
This is the operational readiness problem in its purest form, and it belongs in the same family as the enactment gap: an approved plan that cannot be performed, because the people who must perform it are not yet able to.
What route competency actually costs
Take a freight operator with roughly 10,000 route kilometres and around 2,000 drivers — a realistic profile for a large Australian heavy-haul business, and one that has been doubling in size over a decade, largely through acquisition.
Annual driver intake runs at about 15%, so 300 new drivers a year. Base rate is approximately A$160,000 plus superannuation at 11%, so call it A$165,000 fully loaded. Route competency for a new driver takes at least twelve months, and on some heavy-haul corridors closer to twenty-four. For complex routes a driver new to the road may need to run it up to thirty-five times before the nuances are learned.
That intake alone, carried for twelve months of route competency training, is roughly **A$49.5 million a year**.
Then the recurring half, which is the part people forget. Complex route qualifications lapse after twelve months and must be re-signed. Classroom theory competence holds for thirty-six months, and even that is exposed to seasonal variation. If 10% of the existing 2,000 skilled drivers need new or renewed route qualifications in a given year, that is 200 drivers on a twelve-month cycle — another **A$40 million a year**.
Add the administrative and rostering overhead conservatively at A$1 million, and the total hidden cost of route qualification for one operator is in the order of **A$100 million a year, or A$500 million across five years**.
Two things about that figure matter more than its precision. The first is that it is opex, so it is diffuse, budgeted departmentally and never aggregated. The second is demographic: the average train driver is now over 57, and attrition through retirement is the dominant workforce risk in the sector. The intake percentage is not going down.
Why it gates the programme, not just the training budget
For a project director the opex number is interesting. The schedule consequence is the one that should stop them.
On a new or upgraded route, services cannot run until enough crew are route qualified. In one case a corridor had 180 drivers qualified north-south, against a rule of thumb that a minimum of 60% of the driver pool must be route qualified before you can credibly plan to run services at all. Getting 360 drivers through physical site familiarisation takes twelve to fourteen weeks.
That is twelve to fourteen weeks the construction team loses from its own programme, because the two activities compete for the same asset: the track. And driver rostering is locked six weeks in advance in every operator I have worked with, so the familiarisation window cannot be compressed by goodwill on the day. It has to be planned a quarter out, or it does not happen.
Put those together and the sequence is unforgiving. Commissioning cannot start until crew are qualified. Crew cannot be qualified until they have the track. The construction team cannot release the track without giving up programme. And nobody can flex the roster inside six weeks. A day of that is worth at least A$10,000 in programme terms, and usually a great deal more.
This is why operational readiness is not a training line item. It is a critical path activity that happens to be staffed by the training department.
Why drone video does not work
The obvious shortcut is to film the route and show it to drivers. Operators try it, and it does not produce competency. The reason is specific and worth understanding, because it explains what any substitute for physical familiarisation actually has to do.
A train driver’s eye position is roughly two metres above the railhead, and it is usually offset from the track centreline rather than sitting on it — and both of those vary with the rolling stock being driven. Route knowledge is the recognition of cues in a particular geometry: where a signal first becomes visible, what occludes it, at what point a gradient or curve changes what you can see, how a structure reads at approach speed from that seat.
Drone footage is captured from none of those positions. It produces a beautiful record of the corridor and a useless one for sighting. A driver cannot learn where a signal appears from a viewpoint no driver will ever occupy. Get the eye position wrong and you have made a documentary, not a qualification.
What changes when crew are engaged before they reach the track
Where operators have moved route induction and qualification into survey-accurate simulation at the correct eye position, the pattern reported to us is consistent: around 95% of drivers complete route induction with no physical driving induction at all, the training burden falls by A$1 million to A$2 million on a single corridor, and — the part a project director cares about — the twelve to fourteen weeks go back into the construction programme.
I should be straight about the status of those figures. They are operator-reported outcomes from our own engagements, not an independent study, and there is no controlled comparison behind them. What can be said with more confidence comes from the wider evidence base, and it is more useful anyway.
Demonstration transmits the shape of a sequence far better than description transmits its result: an effect size of d = 0.77 on movement form against d = 0.17 on movement outcome, and the largest effects of all — d = 1.62 — on serial, multi-step tasks. Route knowledge is a serial task. Simulation-based training transfers to real performance at attenuating but real effect sizes, around 1.1 measured in the simulator, 0.8 on behaviour in the real setting, 0.50 on outcomes. And passive watching is not the intervention: across 95 studies and 20,991 workers, knowledge gains ran d = 0.55 for lectures, pamphlets and video against d = 1.46 for behavioural modelling with practice and dialogue.
That last finding is the one that separates a rehearsal from a video. The driver has to do something with what they saw, and somebody has to be able to show who has demonstrated what.
What a project director should ask
Four questions, and none of them are training questions.
**How many crew must be route qualified before we can run services, and how many are today?** If nobody can answer this in a number, operational readiness is not being managed, it is being hoped for.
**How many weeks of track access does crew familiarisation need, and whose programme are those weeks coming out of?** They are coming out of somebody’s. Better to know whose before the argument happens at week forty.
**What is the rostering lead time, and is the familiarisation window inside it?** Six weeks is typical. A plan that assumes crew can be redeployed inside that is not a plan.
**Who has demonstrated route competency, at the level of the individual, and how would we prove it to a regulator?** A signed matrix records that a process ran. It does not evidence that a person can perform.
Frequently asked questions
What is route competency, and why does it expire?
Route competency is a driver’s certified knowledge of a specific route — its signals, sighting distances, gradients, speed restrictions and hazards — from the driving position of the rolling stock they will operate. Complex route qualifications typically lapse after twelve months and must be re-signed. Classroom theory competence commonly holds for thirty-six months. The expiry is what makes it a recurring cost rather than a one-off.
Why does route familiarisation sit on the critical path?
Because services cannot run until a sufficient proportion of the driver pool is qualified, and qualification traditionally requires access to the track. On a live corridor that access competes directly with construction. Twelve to fourteen weeks is a normal duration for getting a few hundred drivers through, and with rosters locked six weeks ahead the window cannot be recovered late.
How much does route qualification actually cost an operator?
For a large freight operator with around 2,000 drivers and a 15% annual intake, the loaded cost of the intake’s twelve-month route competency is roughly A$49.5 million a year, with renewals for existing drivers adding around A$40 million, and administration on top. The order of magnitude is A$100 million a year — around A$500 million over five years — and it is almost never aggregated into a single line anyone is accountable for.
Can drone or cab-mounted video replace route familiarisation?
Drone video cannot, because it is captured from a viewpoint no driver occupies. A driver’s eye position is around two metres above the railhead and usually offset from the centreline, and both vary by rolling stock. Sighting is the recognition of cues in a specific geometry, so a wrong viewpoint teaches the wrong cues. Correct eye position is the minimum condition for any substitute to be worth running.
What is the workforce demographic risk?
The average train driver is now over 57. Attrition through retirement, not recruitment volume, is the dominant risk, and it means the annual intake percentage — and therefore the route competency bill — is structurally rising rather than falling.
Figures for cost, duration and simulation outcomes in this article are drawn from operator engagements and are reported as such rather than as independent findings. The behavioural effect sizes cited are from published meta-analyses: Ashford, Bennett and Davids (2006) on observational modelling; Burke et al. (2006) on safety training method and engagement; Cook et al. (2011, 2012) on simulation transfer.
