Pedestrians are hurt by forklifts more often than drivers are. NIOSH’s long-term studies, as summarized in the material-handling literature, attribute about 20% of forklift accident victims to workers struck while on foot, and another 16% to workers crushed by a forklift, roughly 36% of all forklift-related injuries before overturns, which happen to drivers, are even counted. In the federal severe-injury dataset this site publishes, 5,584 forklift-involved reports across 2015–2025, the dominant patterns are struck-by and caught-between events, both of which are pedestrian mechanisms. This guide turns those numbers into a pedestrian protection program: separate the walker from the truck, make both visible, train both, and add detection where separation cannot reach. It sells none of these systems.
Why pedestrians get hurt
The reasons cluster into four predictable situations. People focused on a task, a pick, a pack, a count, step off a workstation or out of an aisle without looking, as the workstation guidance from Cisco-Eagle describes. Forklifts emerge from blind corners, trailer interiors, and rack aisles faster than a walker expects. Drivers reversing or carrying tall loads cannot see the pedestrian at all, the visibility problem mapped in the forklift blind spots guide. And shared aisles put walkers and trucks in the same corridor by design, which means the encounter rate is a planning choice, not a fact of life. Each situation has a named control, and none of them requires a capital project on day one.
Separation first: the layout layer
The strongest pedestrian protection is physical separation, because it works even when both parties make mistakes. The controls, in order of permanence, are designated pedestrian walkways, guardrails and bollards at work cells and aisle ends, gates at high-traffic doorways, and painted go/no-go floor markings in lower traffic zones. Cisco-Eagle’s guidance on work cells is direct: the pedestrian can easily back out of a work area and into the path of forklifts emerging from a narrow pallet-rack aisle, and the issue is best fixed with a fixed guardrail parallel to the aisle, with sensors, gates, and mirrors as complements rather than replacements. Wooden rails are not strong enough for certain forklift impacts, so barrier strength should match the trucks that travel nearby. Where work must happen near traffic, move the work behind a barrier or into a defined cell before considering any electronic aid.
Visibility: painted edges, mirrors, and predictable rules
Where separation is incomplete, visibility becomes the control. OSHA’s eTool for powered industrial trucks recommends painting the edges of loading docks and marked travel paths to improve visibility, keeping working surfaces clear and clean, and watching for tail swing, because a pedestrian standing at a safe distance from the front of a turning truck can still be struck by the counterweight end. Mirrors at blind corners and on trucks are cheap and effective, with the caveat that they depend on someone looking at the right moment. Rules make visibility predictable: drivers sound the horn at cross aisles and obstructed locations, make eye contact before passing, and give pedestrians the right of way, while pedestrians treat every truck as if the operator cannot see them, especially around reversing trucks, tail swings, and emerging aisles. The safety rules guide turns those expectations into a reviewable list.
Training and the near-miss log
A pedestrian program without training is a set of paint stripes. Operators need refresher triggers defined by the powered-industrial-truck standard, which the plain-English guide to 1910.178 covers, and walkers need the same hazard awareness: exclusion zones, high-visibility clothing where the site requires it, and the rule that a forklift cannot steer or brake like a car. The near-miss log is the program’s instrument panel. Because voluntary near-miss reporting captures only a small fraction of events, as the telemetry guide notes, the log should record at least the near-misses people admit to, and review them monthly for hotspots: shift changes, specific aisles, specific doors. A flat or empty log is a reporting problem, not a safety success.
The detection layer: where technology fits
Detection systems are the backup layer, not the primary one. Pedestrian detection systems warn or intervene when a person enters a danger zone, with the tag-based families on the proximity detection page suited to sites that can control who carries a tag and the camera-AI families suited to visitor-heavy operations where nobody wears anything. Approach lights, the blue light family, mark the danger zone on the floor cheaply. The systems index compares the families side by side, and the choosing a system guide walks through scoring your exposure before any vendor conversation. One caution applies across the layer: a detection system that alerts constantly becomes noise, the alarm fatigue failure covered in the alarm fatigue guide, so the alert rate per shift belongs on the monthly review like any other metric.
The defensible order
The sequence that survives scrutiny is: score the exposure with the Risk Scorer, which turns your routes, intersections, and near-miss log into a 0–100 score and a priority order in about two minutes; separate people from trucks with layout controls where the score points at shared aisles; make both sides visible with markings, mirrors, and rules; train and review the near-miss log on a named cadence; and only then add detection where separation and visibility cannot reach. That order, applied site by site, is how the decade of data on the methodology page suggests the flat national injury curve gets bent, and your state’s injury series provides the local denominator for the budget meeting. Run the Risk Scorer first, compare the detection families on the systems index, and when the scope is pinned, get comparable numbers through the quote form.