Industrial manufacturing safety: reducing machine-on-person incidents

Why machine-on-person risk is so hard to eliminate

Manufacturing environments are designed for production. Machines, vehicles, and people operate in the same space because that's what the process requires. The challenge isn't that machine-on-person risk exists; it's that it's built into the way manufacturing works.

Vehicles sharing space with people on foot

On most manufacturing floors, forklifts, tow tractors, and automated guided vehicles (AGVs) share the same space as workers moving between workstations, maintenance areas, and break rooms. OSHA estimates between 35,000 and 62,000 forklift-related injuries occur annually in the US, with 36% of forklift fatalities involving pedestrians. Forklifts were the source of 84 work-related deaths in 2024 and over 25,000 DART cases.

In manufacturing, the risk is amplified by the complexity of the floor layout. Production lines, staging areas, raw materials storage, finished goods, and shipping zones all create traffic patterns that bring vehicles and pedestrians into proximity repeatedly throughout every shift. Blind corners, noise from production equipment, and the focused attention workers give to their tasks all reduce awareness of approaching vehicles.

Machinery and moving parts

Workers who operate or work near manufacturing machinery face 18,000 amputations, lacerations, crush injuries, and abrasions per year in the US, along with more than 800 deaths. Machine guarding is consistently one of OSHA's most cited standards, and the manufacturing sector led all industries with 937 forklift-related citations in 2024, totalling over $2.7 million in penalties.

Many of the most serious machinery injuries occur not during normal production but during cleaning, maintenance, changeovers, and jam-clearing, the moments when workers interact directly with equipment that's usually guarded or automated during regular operation. Lockout/tagout failures remain a persistent contributor to these incidents.

The space between procedures and reality

Every manufacturing operation has safety procedures: traffic management plans, pedestrian walkways, exclusion zones, lockout/tagout protocols, and PPE requirements. The gap between what procedures say should happen and what actually happens across three shifts, seven days a week, is where most incidents occur.

A worker takes a shortcut through a forklift lane to save 30 seconds. A pedestrian walkway gets blocked by staged materials and people route around it. An exclusion zone is gradually encroached upon as production demand increases. These aren't failures of awareness. They're the slow erosion of compliance that happens in every busy manufacturing environment, and they're almost impossible to detect through periodic walk-throughs and audits alone.

The visibility gap on the manufacturing floor

The fundamental challenge is that near misses, which vastly outnumber actual incidents, are rarely captured. A forklift passes too close to a worker at an intersection, but nobody is hurt and nobody reports it. The same near miss happens three times a week at the same intersection, building a pattern that predicts a future incident, but without continuous monitoring, that pattern remains invisible.

Safety teams end up making decisions based on the incidents that were serious enough to be reported, rather than the leading indicators that would allow them to intervene before someone gets hurt.

How computer vision AI changes the equation

Computer vision AI uses your existing CCTV cameras to continuously monitor the manufacturing floor for the machine-on-person events that drive the most serious injuries.

What it detects

The most critical detections in a manufacturing environment include pedestrian-vehicle near misses (forklifts, tow tractors, and other mobile equipment), exclusion zone breaches near machinery, production lines, or restricted areas, vehicle speed violations on the factory floor, and patterns in how people and vehicles move through the facility across shifts and operational phases.

The system doesn't need to cover every camera in the plant. It focuses on the highest-risk zones: the intersections where vehicle and pedestrian traffic cross, the areas around heavy equipment, the dock and dispatch zones, and any blind spots where visibility is limited.

Revealing patterns that predict incidents

The heatmap and reporting tools aggregate safety events over time, showing your team exactly which zones, shifts, and operational patterns generate the most risk.

Many inviol customers discover that their plant layouts or operational schedules are creating more risk than they expected. A particular intersection near the raw materials staging area generates three times the near-miss rate of anywhere else in the plant. Or vehicle speed violations increase during the final two hours of each shift as throughput pressure builds. Or a pedestrian crossing point that was safe when the plant ran one production line becomes dangerous when a second line is added and traffic patterns change.

These insights lead to targeted, practical changes: repositioning a pedestrian walkway, adding a physical barrier at a blind corner, adjusting a vehicle route, or staggering a material delivery schedule. Changes that reduce risk and often improve flow and throughput at the same time.

From detection to coaching

Every captured safety event becomes the starting point for a coaching conversation. Video clips with faces blurred for privacy are shared with the team: what happened, what could have gone wrong, and what should change.

In a manufacturing environment, this coaching approach is especially valuable because it bridges the gap between procedures and reality. Instead of telling workers what the procedure is, you're showing them what actually happens on the floor and having a conversation about how to improve it. That's a fundamentally different dynamic from traditional compliance-based training, and it produces lasting behaviour change.

inviol customers typically see an average 67% reduction in risk and a 42% reduction in incidents across their sites, with a 61% reduction in machine-on-plant incidents specifically.

Compliance and your obligations

Under New Zealand's Health and Safety at Work Act (HSWA), PCBUs must eliminate or minimise risks so far as is reasonably practicable. For a manufacturing operation, that includes managing vehicle-pedestrian interactions, machinery access, and the specific hazards created by production processes. A system that continuously monitors high-risk zones and documents how safety events are identified, coached, and resolved provides strong evidence of meeting your duty of care.

In Australia, Safe Work Australia reports that manufacturing fatalities in 2024 were 44% lower than the five-year average, a positive trend that demonstrates targeted safety improvements can make a measurable difference. However, the industry remains among the top six for serious workers' compensation claims, and regulators expect continued investment in proactive risk management.

In the US, OSHA has driven worker injuries down from 10.9 per 100 workers in 1972 to 2.4 in 2023, but machine guarding, powered industrial trucks, and lockout/tagout remain among the most frequently cited standards in manufacturing. The 2023 launch of OSHA's warehouse and distribution NEP also covers manufacturing operations with high injury rates.

Getting started

Start with the intersections where vehicle and pedestrian traffic most frequently overlap, and the zones around your heaviest equipment. These are the areas where machine-on-person incidents are most likely to occur, and where the gap between procedures and reality is widest.

Computer vision AI works with your existing CCTV cameras, processes data on-premise for privacy, and gives your safety team the leading indicators they need to prevent the next incident.

Book a demo and we'll show you how it works for manufacturing operations like yours.

Frequently Asked Questions

What are the most common machine-on-person incidents in manufacturing?

The most common incidents include workers being struck by forklifts or mobile equipment, caught-in or caught-between events involving machinery, contact with moving parts during maintenance or changeovers, and pedestrians entering exclusion zones near heavy equipment. Contact with objects and equipment is the third leading cause of workplace death in the US.

How many injuries do forklifts cause in manufacturing?

OSHA estimates between 35,000 and 62,000 forklift-related injuries occur annually in the US across all industries, with manufacturing leading all sectors in forklift-related citations. Forklifts caused 84 work-related deaths in 2024, and 36% of forklift fatalities involve pedestrians.

How can AI reduce machine-on-person incidents?

Computer vision AI uses existing CCTV cameras to continuously monitor the factory floor for safety events like pedestrian-vehicle near misses, exclusion zone breaches, and speed violations. It captures the near misses that go unreported and generates heatmaps that reveal risk patterns, enabling targeted process improvements. inviol customers see a 61% reduction in machine-on-plant incidents.

Why do near misses matter more than incident reports?

Near misses vastly outnumber actual incidents and represent the leading indicators that predict future injuries. In manufacturing, a forklift near miss at the same intersection three times a week is a clear signal that an incident is likely. Without continuous monitoring, these patterns remain invisible to safety teams making decisions based only on reported incidents.

How does coaching improve safety in manufacturing?

When a safety event is detected, the video clip is captured with faces blurred for privacy and used for a team coaching conversation. This bridges the gap between written procedures and what actually happens on the floor, showing teams real events and discussing how to improve. This approach produces lasting behaviour change and inviol customers typically see a 67% reduction in risk.