Vehicle incidents are 42% of work deaths. The defences that fail are the same ones every time.

Vehicle incidents are 42% of work deaths. The defences that fail are the same ones every time.188 workers died at work in 2024 and vehicle incidents caused 42% of those deaths. What a run of vehicle investigations usually finds in the Absent/Failed Defences family, and how to read across them.Safe Work Australia's Key WHS Statistics 2025 record 188 worker deaths in 2024. Vehicle incidents were 42% of them, the largest single mechanism for as long as the series has run. Falls from height were next at 13%. Six industries carried 80% of the deaths: agriculture, forestry and fishing; public administration and safety; transport, postal and warehousing; manufacturing; health care and social assistance; and construction.If you run a fleet, a mine, a farm or a site with mobile plant, you've investigated a vehicle incident. Probably several. Read in isolation, each one has its own story. Read as a set, they tend to say the same thing about the Absent/Failed Defences (DF) family, and that's where the next one is coming from.What the DF findings usually say

Separation that relied on behaviour. Pedestrian and vehicle interaction controlled by a rule and a line on the ground, with no physical barrier where the two mix

Fatigue management that stopped at the roster. Hours were within limits, but the roster and the drive home weren't looked at together

Vehicle fitness that was a signature. The pre-start was signed every day. The brake fault was in the maintenance system for three weeks

Traffic management plans that described a site that had changed. Haul road realigned, plan not updated, inductions still taught the old layout

Communication that assumed line of sight. Radio protocols existed. The reversing excavator and the light vehicle were on different channels

Why the person-level finding keeps winningVehicle incidents are the category where investigations most often stop at Individual/Team Actions (IT). The driver was speeding, the spotter was out of position, the operator didn't check. Those findings are frequently true. They are also the least useful finding an investigation can produce, because the next driver inherits the same haul road, the same roster and the same radio channels.The discipline is to ask, for every IT finding, which defence should have caught it. A driver who was speeding on a haul road with no speed control, no proximity detection and a road design that invited it is an IT finding sitting on top of three DF findings. The DF findings are where the recommendations go. Task/Environmental Conditions (TE) usually hold the rest: visibility, road surface, the fatigue codes under HF.Reading across the setPull every vehicle investigation from the last two years and list the DF findings side by side. If the same defence appears three times under slightly different wording, that is one failed control, not three incidents. Name it once, put one corrective action against it at the engineering level of the hierarchy, and track that action to verified closure. The reports already hold this. What's usually missing is the consistent naming that lets you see it.SafetyPulse classifies every factor against the ICAM families and codes, so a search for failed defences across vehicle investigations returns a list, not a reading exercise. Where an organisation runs bowties for its vehicle interaction risk, each investigation's barrier findings update the bowtie, and the barrier that keeps failing shows as failing. The investigation you run next month is better for the ten you ran before it, if the ten can be read together. Features · How it works · Methodology · Security · Resources · Book a demo SafetyPulse by ICAM Australia Pty Ltd. ICAM training and methodology.