Forklift and Pedestrian Safety with Cameras in Indian Factories

Camera-based forklift safety uses computer vision to watch the space around a truck — or a fixed floor zone — and warns the moment a person and a forklift share dangerous ground. Forklift-mounted cameras cut blind-spot strikes; fixed overhead cameras enforce lane and keep-clear zones. Both catch pedestrian intrusion in real time, but neither sees through racking or replaces physical separation.

Pedestrians struck by forklifts are the single largest category of lift-truck deaths. In US data, forklifts were the source of 84 work-related deaths in 2024, and roughly 36% of forklift fatalities involve a pedestrian — even though pedestrians are only about a fifth of all forklift incidents (OSHA / NSC Injury Facts). India does not publish an equivalent national forklift-pedestrian fatality series, so treat US figures as directional, not local — but the physics of a 3-tonne truck and a walking worker are identical in a Pune auto-components shed or a Bhiwandi FMCG warehouse.

This article covers the two camera approaches, what each realistically catches and misses, an honest comparison with UWB/tag systems, indicative Indian pricing, and where this sits under the Factories Act, 1948.

Two camera approaches — they solve different problems

"Camera safety for forklifts" means two quite different deployments. Owners routinely conflate them, then buy the wrong one.

1. Forklift-mounted AI cameras (moving envelope)

A camera and a small AI box ride on the truck, watching the operator's blind zones — behind the mast, around a tall load, at corners. AI distinguishes a person from a rack leg or a pallet, and fires tiered alerts: an audible/visual warning when someone enters the warning zone — Indian vendor listings put the adjustable range at roughly 1 to 6.5 m (Pedestrian Alert System listing) — escalating to auto-deceleration or braking on some systems as the gap closes (WTSAFE). This travels with the hazard, so it protects wherever the truck goes.

2. Fixed zone-intrusion cameras (virtual lanes on the floor)

Overhead or wall-mounted cameras — often the CCTV you already run — with analytics that draw virtual forklift lanes and pedestrian keep-clear zones on the floor. When a person steps into a live forklift aisle, or a truck crosses onto a walkway, the system logs it and alerts a supervisor. This is the same zone-and-path logic used for restricted-area intrusion, applied to traffic separation. It enforces your floor plan rather than one truck's bubble — and it produces a timestamped compliance record, not just a live beep in a cab.

Most serious plants end up wanting both: mounted cameras for the operator's blind spots, fixed cameras for lane discipline and evidence.

What cameras catch — and what they miss

Vision-based safety is genuinely useful, but selling it as foolproof gets people hurt. Be clear-eyed.

Catches well: - A pedestrian walking, standing, or crouching in a clear line of sight, day or reasonably lit night. - A truck straying out of its lane or a person entering a marked keep-clear zone (fixed cameras). - Multiple people at once; distinguishing humans from static objects to cut nuisance alarms (Detect Technologies on vision AI). - Blind-spot coverage behind a tall load where mirrors fail (mounted cameras).

Misses or degrades on: - Occlusion. A person fully hidden behind racking, a stacked pallet, or another truck is not visible to a camera — this is the core limit in dense Indian warehouses with narrow aisles. - Poor conditions. Heavy dust, smoke, fog, glare, or genuine darkness degrade vision detection; harsh-environment reliability is where tag/radar systems claim an edge (ELOKON camera-vs-UWB). - Sightline gaps. Fixed cameras only cover aisles you actually framed; a lane with no camera has no enforcement. - Alert fatigue. Any system that cries wolf trains operators to ignore it — tuning zones to your real traffic matters more than the spec sheet.

Camera vs UWB tag vs physical separation

There is no single winner; the right mix depends on aisle density, dust, and whether your workers will reliably wear a tag.

The honest read: cameras and tags are complementary, and neither replaces barriers, marked lanes and speed control — which regulators expect as the baseline anyway.

Indicative costs in India (mid-2026)

Public pricing is thin, so treat every figure as an indicative estimate, not a quote. One Indian single-unit listing for an AI camera-based forklift pedestrian-alert unit sits at ₹40,000 per forklift (Faridabad — IndiaMART, mid-2026); a separate Indian listing for such a unit quotes an adjustable detection range of roughly 1–6.5 m with person-vs-object discrimination (Pedestrian Alert System). Multi-camera 360° kits and auto-braking integration cost more.

Fixed zone-intrusion analytics is usually priced per camera-channel per month as software, on top of your existing cameras and a PoE/networking budget — a different cost model from the one-time per-truck box. Get written quotes for your actual truck count and aisle count; safety-only scopes sit far below full multi-use-case ones.

Where a plant head should start

Don't wire the whole plant. Follow the risk:

• Map your worst conflict points first — dock aisles, blind corners, the walkway crossing the main forklift artery. That is where the first mounted unit and the first fixed zone go.
• Reuse existing floor cameras for lane/keep-clear zones where sightlines are already clear; add cameras only where the aisle boundary is currently unframed.
• Keep the physical baseline — painted lanes, barriers at pedestrian crossings, speed limits. Cameras supervise that discipline; they do not substitute for it.
• Tune before you trust. Set zones to real traffic and cut false alarms early, or operators will tune the system out.

Deciding which aisles need a camera, from where, to frame a forklift lane without a blind spot is exactly the survey headache Mama removes: you record a short phone walkthrough of the floor, and it reads the space — aisles, crossings, racking occlusion, sightlines — then returns a floor plan plus a camera-placement plan showing which forklift lanes and pedestrian zones each camera can actually cover, before you buy a single unit.

Factories Act, 1948 — where this fits

Internal transport safety in India rides on the occupier's general duty, not a forklift-specific rule. Section 7A of the Factories Act, 1948 (introduced by the 1987 amendment) requires every occupier to ensure, so far as is reasonably practicable, the health and safety of all workers — expressly including safe systems of work and safe arrangements for handling, storage and movement (India Code, §7A; full Act text). Separating pedestrians from powered trucks is squarely a "safe system of work," and a timestamped intrusion log is evidence you are managing it. State Factory Rules and your inspectorate add specifics; the technical advisory arm is DGFASLI (now DGOSH), which advises on factory OSH. Forklift terminology and classes follow BIS IS 4660 (Powered Industrial Trucks, aligned to ISO 5053). Analytics does not discharge the duty — guarding, lanes and training do — but it gives you a continuous record that your controls are actually followed.

FAQ

Do forklift safety cameras stop accidents or just record them? Both, depending on type. Forklift-mounted AI cameras alert the operator in real time and, on some systems, trigger auto-deceleration or braking — actively preventing strikes. Fixed zone-intrusion cameras mainly alert a supervisor and log breaches, so they enforce lane discipline and build evidence rather than braking a truck.

Will a camera see a worker hidden behind racking? No. Occlusion is the core limit of any vision system — a person fully blocked by racking, pallets or another truck is invisible to the camera. In dense, narrow-aisle Indian warehouses this matters, which is why UWB/RFID tags (which see through obstacles) are used alongside cameras, and why painted lanes and barriers stay essential.

Camera or UWB tag system — which is better for an Indian warehouse? Neither wins outright. Cameras need no worn tag and give visual context but fail on occlusion, dust and darkness; UWB tags see through racking and work in dust but fail if the tag isn't worn or charged. Dusty, dark, high-rack sites often combine both. Match the choice to your aisle density and whether workers will reliably wear tags.

What does forklift pedestrian detection cost in India? Indicative only: Indian listings for an AI camera-based forklift pedestrian-alert unit run around ₹40,000 per forklift (mid-2026), with 360° multi-camera and auto-braking kits costing more. Fixed zone-intrusion analytics is priced per camera-channel per month on top of existing cameras. Always get written quotes for your truck and aisle count.

Is forklift-pedestrian separation legally required in Indian factories? There is no forklift-specific section, but Section 7A of the Factories Act, 1948 makes the occupier responsible for a safe system of work "so far as is reasonably practicable" — which covers separating pedestrians from powered trucks. State Factory Rules add specifics; DGFASLI/DGOSH is the technical advisory body. Cameras help evidence compliance but do not replace barriers, lanes and training.