Opens a sample AR Experience
works on any Smartphone
Buyers in high-risk industries are usually choosing between about ten different kinds of supplier, and it is rarely obvious that they do different jobs. Some teach people about the work. Some record the work after it has happened. Very few do anything at the moment the work is being done.
The most useful way we have found to describe the market is on two questions rather than a feature list:
Where does it live? In the training room, or at the point of work?
How deep does the model go? Does it store documents and forms, or does it compute what happens?
Plot the categories on those two questions and the market sorts itself into four groups.
Categories of supplier, not companies. Positions describe what each category is designed to do, as we understand the market in August 2026.
The categories on the left know the training buyer and stop at the classroom door. The categories on the right sit inside operations and begin only once a worker is already assumed to be competent.
We built Work Intelligence to work in both halves, because the events our customers care most about tend to happen in the space between them.
It depends entirely on what the buyer is trying to solve, which is why we answer in categories rather than names.
If you want people to practise judgment in hazardous situations, you are comparing us with immersive safety content. If you want high-fidelity process behaviour for control-room operators, you are comparing us with operator training simulators. If you want digital work instructions on a tablet, you are comparing us with connected worker platforms. If you want the qualification record kept straight, you are comparing us with competency management.
Those are four different purchases with four different buyers. Most organisations end up owning several of them, and there is nothing wrong with that. What we ask is that you know which one you are buying.
We do not publish claims about individual competitors’ products. Their capabilities change, we cannot verify them from the outside, and we would rather you tested ours than took our word about somebody else’s.
No, and the distinction is worth understanding because it changes what the data at the end of it means.
Most immersive safety content is authored: an artist builds a scene, a writer scripts the situation, and each choice leads to an outcome that a person designed in advance. It is very good at what it is for — rehearsing a judgment call until the reflex is there.
Work Intelligence is built the other way round. We generate a model of the equipment, in which components carry their own function and behaviour, and a physics layer computes what happens when the machine is operated. Nobody has to have anticipated the sequence for the model to respond to it.
The practical difference shows up in two places: whether your own equipment can appear in the scenario at all, and whether the system can answer why something failed rather than only that it did.
Operator training simulators are the deepest models in this market. They are built for control-room operators, coupled to the control system, and they represent process behaviour to a level of fidelity we do not claim to match.
They are also, by design, a control-room product. They serve a small number of panel operators, they take substantial engineering effort to build and maintain, and they do not follow a technician out to the equipment.
Our physics layer is lighter, and it runs on a phone or a tablet for the whole workforce. If your problem is panel operator fidelity, a simulator vendor is the right answer. If your problem is the several thousand people who never sit at that panel, it is not.
Connected worker platforms replaced paper with a screen and connected it to the maintenance system, which was a genuine improvement. What they capture is what the worker recorded: a step marked complete, a value typed in, a signature.
Work Intelligence adds a different kind of capture. The system recognises the equipment in front of the worker, guides the step against a model of that equipment, and records the difference between the guidance and what actually happened.
Many customers will want both. We are designed to sit alongside, not to replace the system your operations team already trusts.
An LMS answers whether someone completed a course. That is a real question and it needs a system.
It is not the same question as whether the person can perform the task on the equipment you actually operate. We are built for the second question, and we integrate with the system that answers the first.
Competency management systems hold the record: who is qualified for what, assessed by whom, valid until when. They are the system of record and they should stay the system of record.
What they hold is a judgement that somebody else made somewhere else. The gap most organisations run into is producing the evidence behind that judgement at the scale their regulator or their auditor expects.
Some regimes are explicit about this. In the United States, the pipeline operator qualification rules at 49 CFR 192 Subpart N and 195 Subpart G set out which evaluation methods an operator may use, and simulation is among them. Read your own rule rather than our summary of it – and then ask which of your systems produces each kind of evidence it permits.
No, and we would advise against anyone who says they do. Control of work is a controlled system, it is audited, and it carries authority. We do not issue permits and we do not authorise work.
What we do is help the person in front of the paperwork understand it, prepare it correctly and learn it. The authority stays exactly where your management system puts it.
Usually not, although the word is doing a lot of work in both cases.
Engineering digital twins are built for engineers: accurate geometry, asset data, and a great deal of contextualised information about the plant. They are excellent for reliability and design and they are not built for a human being to learn or perform a task inside.
Ours is built for the worker. Components carry what they are and what they do, procedures run on them, and the model is generated from photographs or a P&ID rather than surveyed. If you already have an engineering twin, that is a head start, not a conflict.
No. Accreditation, assessor authority and the certificate itself sit with accredited providers and standards bodies, and none of that is something we offer or want to offer.
Where we help is the practice and the evidence in between: repeatable rehearsal on the equipment your people actually operate, without moving everybody to a training centre first. Several providers in this category are partners rather than competitors for exactly that reason.
These four separate the categories faster than any feature comparison, and we are happy to be judged by our own answers to them.
A fair page has to answer this one.
Do not take a comparison page – ours or anyone else’s – as evidence. Two tests are worth more than a month of evaluation documents:
Bring your own equipment. Send a photograph of a machine you actually run and ask each vendor to put it into their product, and to tell you how long it takes and what it costs. The answers differ enormously and the difference is structural, not commercial.
Ask for the evidence artefact. Ask each vendor to show you exactly what a regulator, an insurer or an internal auditor would receive at the end. Then decide which of them is producing evidence and which is producing a claim.
