This article was originally published on LinkedIn by Ozgur Ulku, Co-Founder & CEO of Litum, on September 23, 2026.
Every time I hear about another forklift collision on a plant floor, it reinforces something we have believed at Litum for a long time.
The technology to help reduce these encounters already exists. It has existed for years.
That is not the hard part.
Ask anyone who runs an industrial operation whether forklift safety matters and you will not get an argument. The numbers are not controversial either. Forklifts continue to be involved in serious workplace injuries and fatalities, with pedestrians particularly exposed when people and vehicles share the same operating space.
So the question I keep coming back to is not whether sites believe in forklift safety. It is why so many of them still run on training, signage, mirrors, and the hope that an experienced operator will see what is behind them.
I have spent more than two decades in these conversations, in plants and yards across more than 50 countries.
My answer is that safety projects rarely die because someone objected. They die because of everything that has to be true before they can start.
Four obstacles come up most often.
1. The project keeps getting scoped as a building project
A site sets out to improve forklift safety and quickly finds itself pricing something much larger.
Anchors. A facility-wide mesh network. A site survey. A conversation with IT. A capital request that now has to compete with production spend.
The timeline stretches past the budget year and the project moves to the list of things to revisit.
Nothing was rejected. It simply never became small enough to approve.
I want to be fair here. Facility-wide RTLS delivers things a single vehicle cannot: traffic patterns, congestion visibility, movement analysis across an entire operation. That value is real.
But it is a second question. Whether a forklift can sense a pedestrian behind it while reversing should not wait on whether the whole building has been instrumented.
A safety project should be able to begin with the vehicles where the exposure is highest and expand from there, without requiring the site to solve the entire facility on day one.
2. The fleet will not hold still long enough
Very few sites run one forklift brand on one replacement cycle. They run several makes, several models, several vehicle types, some leased and some owned, with equipment cycling in and out as leases turn over.
A safety system tied to one manufacturer, or one that has to be re-specified every time a truck is swapped, quietly stops being deployed.
Coverage goes uneven. The newest trucks get protected. The oldest ones, often working the tightest corners of the plant, do not.
The goal was never a uniform fleet. It is a uniform safety standard across a fleet that will never be uniform.
And that safety investment should not have to be reconsidered every time a lease expires or a different truck enters the fleet.
3. Wearables put the burden on the wrong people
This one shows up after go-live, which is why it gets underestimated so often.
If detection depends on every person carrying a tag, then safety depends on distribution, charging, pairing, replacement and enforcement. It depends on contractors and visitors being equipped on day one. It depends on nobody leaving a tag in a locker.
There is also a part that rarely makes it into the business case. Asking every worker to wear a device so that vehicles can detect them can invite a conversation about surveillance, and that conversation can end up defining how the entire program is received.
Safety systems work best when the floor trusts them. That is not a soft consideration. It is an adoption requirement. That means minimizing unnecessary demands on workers, but also making sure the intervention itself is relevant enough that operators continue to respond to it.
Cameras remove the wearable requirement, but they introduce a different set of considerations. Their effectiveness depends on maintaining a clear field of view in the conditions where vehicles actually operate, including dust, changing light and busy visual environments. They can also introduce questions around image processing, storage and employee privacy that organizations may need IT, legal and HR to address.
None of those considerations make cameras unsuitable by definition. They simply add another layer to a project that may have started with a much narrower objective: helping a forklift detect and respond to risk around the vehicle.
4. Meanwhile, the events that matter go unrecorded
Underneath all of this sits a quieter problem, and it is the one that troubles me most.
On most floors, the near miss that made an operator’s stomach drop at 6:40 in the morning is not a record. It is a story, told once, forgotten by lunch, unless reported manually.
The same gap runs through the rest of the operation. Nobody can say with certainty which named operator was in which truck on which shift. Pre-operation inspections often live on clipboards, and a reported defect depends on a supervisor noticing the sheet. Utilization gets discussed in impressions rather than in active time and idle time, which is a difficult place to be when a lease comes up for renewal.
Then a collision happens, and a lot of people go looking for data that was never captured.
A warning in the moment matters. But the value grows when the event does not disappear once the warning stops.
When an organization can see where and when proximity events recur, connect activity to the vehicle and operator, digitize inspection results and understand how equipment is actually being used, safety becomes something the operation can manage rather than simply react to.
Start on the vehicle
Taken together, those obstacles suggest a fairly simple design brief.
A practical forklift safety system should work with the fleet already on the floor. It should be possible to deploy without turning the entire facility into a technology project. It should not depend on every pedestrian carrying a device. Alerts should reflect the conditions of the operation rather than becoming background noise. And when something happens, the event should become useful information rather than disappear with the warning.
That is the thinking our team built PathAware® around: a retrofit system that puts the first layer of intelligence on the forklift instead of on the building.
The core safety layer requires no facility RTLS, fixed infrastructure or mesh network. Pedestrian detection is tagless by default, using AI-enhanced sensing without facial recognition or stored imagery. UWB adds vehicle-to-vehicle awareness beyond the sensor’s line of sight, including in blind spots. Detection and warning zones can be configured to the operating environment, helping sites focus alerts on the interactions that require operator attention.
Because the system is installed on the vehicle, it works across brands, models and vehicle types, on leased and owned equipment alike. Sites can begin with a defined group of forklifts rather than treating deployment as a facility project, then add vehicles progressively as the safety case expands.
You standardize the safety system without standardizing the fleet. And normal changes to that fleet do not require the site to redesign its safety architecture.
PathAware comes in three tiers because not every site needs to solve the same problem on day one.
PathAware® Core focuses on collision awareness on the vehicle. PathAware® IQ adds the connected layer, turning what happens on the floor into safety data the organization can act on and adding capabilities around operator control. PathAware® IQ+ extends that further with digital pre-operation inspections and utilization insights, so the same system can support decisions around both safety and fleet operation.
A site can begin with the trucks in its busiest or highest-risk area, learn from that deployment and expand from there. We have seen this kind of progression repeatedly: starting with a defined safety problem, proving the operating model in the field, and broadening the scope once the case for expansion is clear.
On the economics, the comparison I would run is not simply against the safety budget. It is against the consequence of the incident the investment is intended to prevent.
A serious forklift collision carries costs well beyond the immediate injury or equipment damage: downtime, investigation, lost productivity, workers’ compensation and, in many cases, disruption that reaches far beyond the vehicle involved.
That does not mean every safety investment automatically makes economic sense. It does mean the cost of doing nothing belongs in the calculation.
None of this makes facility-wide RTLS unnecessary. Sites that want traffic analysis, congestion visibility and movement patterns across the whole operation will still want it, and PathAware® is built to extend into that environment when they are ready.
The point is not vehicle intelligence instead of facility intelligence. The safety case simply should not have to wait for the broader infrastructure case.
My argument is only about sequence.
Start where the risk is, on the vehicle. Make the intervention practical. Capture what happens. Learn from it. Expand when the broader business case warrants it.
If your safety project has been sitting on the revisit list for more than a year, I would genuinely like to know which of these four obstacles is holding it there. That feedback has shaped a lot of what we build.
PathAware® is Litum’s forklift safety and telematics product, running on the vehicle rather than the building, with collision awareness, operator control, digital inspections and utilization insights across three tiers.
Follow Litum for more on forklift safety and industrial operations.
