Two near-misses at the blind corner
A 25,000 m2 distribution warehouse runs two shifts with four counterbalance trucks and about 45 pickers. The main aisle is 4.2 m wide and the racking is 8.5 m high. When the racking was re-laid three months ago, the yellow pedestrian walkway was ground off and never repainted. At the despatch end, the aisle meets a doorway with a fast-acting roller door and a blind 90-degree corner; there is no convex mirror. There have been two near-misses in six weeks - one with a temporary picker who stepped out from behind a pallet stack, one with a truck driver walking to his cab - and neither was investigated beyond a note in the shift log. Conveyor noise measures 82 dB(A), so drivers say the horn is useless. The site traffic management plan is dated 2019 and still shows the old layout. A supervisor's proposed fix is a toolbox talk telling pedestrians to make eye contact and telling drivers to sound the horn at every corner.
1. What controls would you put in place before the next shift starts?
Read the worked answer
Before the next shift, the aim is to stop trucks and people sharing the same space at the same time at the point where sight lines fail.
Physically separate them: reinstate a marked pedestrian route with a rigid barrier or bollards alongside the main aisle, and take pedestrians around the despatch end rather than through it. Fit a convex mirror at the blind corner and remove or relocate the pallet stack that blocks the view - nothing should be stored within about a metre of the corner. Fit a pedestrian gate or self-closing barrier at the roller door so people cannot walk straight into truck traffic, and keep the door under control rather than on fast automatic activation.
Control speed and movement: a walking-pace limit at the despatch end with signs, and a rule that no truck reverses around the blind corner without a spotter or a controlled one-way flow. Add flashing beacons, and check whether the horn is actually audible at 82 dB(A) - if not, that is a defect to be fixed, and an audible reversing alarm plus a light is a more reliable warning.
Brief both shifts today, including the agency staff, and stop work at that corner if the mirror and barriers are not in place.
2. Critique the supervisor's proposal using the hierarchy of controls. Why is eye contact an unreliable control?
Read the worked answer
The supervisor's proposal sits almost entirely at the administrative level, with a bit of PPE implied by high-visibility clothing. Administrative controls depend on people behaving correctly every time, and the hierarchy of controls ranks them below eliminating the hazard and below engineering controls for exactly that reason.
Eye contact is unreliable for several specific reasons. The pedestrian cannot see the driver's eyes through a load, a mast, a dirty windscreen or a pillar. The driver's attention is on the load, the aisle, the mirrors and the route ahead, and a truck's blind spots are large, particularly in front of the mast and close to the counterweight. A temporary picker who has never been inducted does not know the convention at all. And when the pedestrian and driver do finally see each other, it may already be too late - a 2.5 t truck at walking pace still travels several metres in the time it takes a person to react.
The horn is equally weak here. At 82 dB(A) of conveyor noise, a horn from behind a rack is easily masked, and if drivers sound it constantly, people stop responding to it. The right order is: remove the interaction by separation, engineer the sight lines and speed, then use training and signs to support those controls, not to replace them.
3. What should be done with the two near-misses, and how would you know the new controls are working?
Read the worked answer
Both near-misses should be investigated properly, even though nobody was hurt. The investigator should get short statements from the picker, the truck driver and the truck driver on foot, walk the route at the time of day the events happened, and record exactly what blocked the line of sight and where each person was standing. There should be no blame for reporting - if the temporary picker is disciplined, the next agency worker will say nothing.
The key question is why two independent events happened at the same corner within six weeks. The common factor is likely to be the layout change, not the individuals, which means the corrective actions belong to the layout and the traffic plan, not to a talk. Re-issue the traffic management plan with the current layout, add temporary pedestrian routes and the corner to the site induction, and give each action an owner and a due date.
To know the controls work, measure rather than assume: observe actual speeds at the corner, count pedestrians using the marked route, check whether the mirror is being used, and review near-miss reports for the corner over the next three months. A rise in reported near-misses after a no-blame launch is usually a sign that reporting is working, not that safety is getting worse.