Most enterprise VR training programmes do not fail in the pilot. They stall right after it. The first site gets headsets, a module and an enthusiastic trainer, and the results look good. Then the second plant asks for the same thing. A third country wants it in its own language, and headquarters wants one set of numbers for all of them.
That is the point where VR stops being a technology project and becomes an operating model. The pressure behind it is real. The World Economic Forum's Future of Jobs Report 2025 estimates that if the world's workforce were 100 people, 59 would need training by 2030. No training team can meet that through more classrooms and more travel.
This guide brings RoT STUDIO's own deployments under one roof, from a single steel plant to an automotive group working across two countries. Each case shows a different level of scale, what changed at that level, and what any organisation can copy from it.
Why Scaling Is the Real Test of VR Training
The research gap between a pilot and a programme is wide. In the Capgemini Research Institute's survey of more than 700 companies, 75 percent of organisations with large-scale VR implementations reported operational benefits above 10 percent. Among small-scale implementers the figure was 59 percent. Scale is where the value actually shows up.
The same research separated a group of “early achievers” from everyone else. Seventy-eight percent of them ran immersive programmes through a centralised governance structure, against 51 percent of the rest. Central ownership, not better headsets, was the clearest difference.
PwC's enterprise study points the same way. It trained more than 1,600 newly promoted managers across 12 US office locations, and the operational lesson was blunt. The number of headsets depends on three questions: how many locations, how many users, and how fast they need training. By early 2020 the firm was supporting 250 headsets across 25 locations, enough to train its entire US workforce in under five weeks.
Four Levels of Scale, Four RoT STUDIO Programmes
Scaling is not one decision. It happens in levels, and each level adds a different problem. The table maps the levels to the programmes covered below.
| Level | What changes | RoT STUDIO case |
|---|---|---|
| 1. One site, live operations | Training must run without stopping production | Tosyalı Holding, iron and steel |
| 2. Many sites, one programme | Content and records must be consistent across factories | Carmeuse Türkiye, six factories |
| 3. Across borders and HQs | Plants in different countries need one method but local ownership | Ford Otosan Romania, Craiova |
| 4. Beyond the workforce | Audiences outside the company, from students to citizens | Mersin Metropolitan Municipality, Gazi University |
Two further programmes, at Fiat TOFAŞ and ErBatu Machine, show the engineering side of the same story. They cover what happens when design data, not just training content, has to move between teams and locations.
Ford Otosan Romania: Scaling Assembly Training Across a Two-Country Group
Ford Otosan is the joint venture of Ford and Koç Holding. In 2022 the European Commission cleared its acquisition of Ford Romania, the company operating the Craiova assembly plant. That move gave Ford Otosan production in two countries. It also gave Craiova a demanding brief: prepare the workforce for new and electric models on the plant's own timeline.
The challenge was a familiar one in automotive. A new model arrives, and the only way to train operators on it is the vehicle itself. One vehicle trains a handful of people at a time, while hundreds need to be ready before the line starts.
Instead of having one vehicle and training a few assembly operators at once, the RoT STUDIO Trainer module has allowed us to train more operators virtually on any stage of the assembly process. It gives us more flexibility, efficiency and focus, while shortening the time to improve the overall assembly process.
The plant chose the RoT STUDIO License with the Trainer module, so its own team could build assembly scenarios from engineering data without writing code. The case study records 210 training sessions in six weeks, and the training team reports fewer critical errors in assembly steps.
The person building that content was not a developer. Girlea Alexandru, Virtual Trainer at Ford Otosan Romania, put it this way. “I am really happy to say that designing my own training scenarios with only CAD knowledge is exciting! No-code platforms are getting stronger and RoT is one of the best so far with its very realistic environment and 1:1 experience as real-life.”
The scaling lesson is ownership. Every future model will arrive with a new training requirement. Because Craiova authors and edits its own modules, that requirement becomes a recurring capability rather than a recurring purchase. The full story is in the Ford Otosan Romania assembly line article.
The group connection runs the other way too. In Türkiye, separately from the Craiova programme, Ford Otosan held an event for the International Day of the Deaf. Participants tried RoT STUDIO's VR Risk Hunt application. The feedback RoT STUDIO received afterwards was that the experience changed how people saw VR.
Seen together, the two activities show one group running VR at two levels at once. Romania uses plant-owned procedural content, while the Türkiye event used ready-made safety awareness training.
Carmeuse Türkiye: Six Factories, One Safety Programme
Carmeuse Türkiye has made occupational health and safety a core priority since 2011, and today it operates across six factories. Its goal was to reduce workplace accidents by changing behaviour and awareness, not only by adding physical controls.
Six factories create the classic multi-site problem. Each site has its own trainers, its own hazards and its own habits. Without a shared system, the same course drifts into six versions, and the head office loses a reliable picture of who has been trained on what.
The programme was built through Customized VR/XR Services, with scenarios drawn from Carmeuse's own operations and delivered through an online platform with centralised management. The Carmeuse Türkiye case study records two awards. One is the Carmeuse Group Innovation Award. The other is the International Lime Association Health & Safety Award, which the association runs with the Vision Zero programme.
Solutions such as customized trainings, the online platform, and centralized management have not only accelerated our processes but also strengthened our employees' safety awareness. Through this collaboration, we have both brought international awards to our country and achieved the goal of zero lost-time accidents.
The scaling lesson is central management with local relevance. Each factory trains in scenarios that look like its own work, while the programme, the records and the standards stay in one place. That is the same pattern Capgemini found among its early achievers.
Tosyalı Holding: Training Inside a Running Steel Plant
In iron and steel, production rarely stops for training. That was the brief at Tosyalı Holding: enable on-site training without halting operations. The work involved the group's training and R&D departments. It also made Tosyalı one of the first to bring VR field training into the Turkish iron and steel industry.
Using the Trainer module, the team built two training modules for two facilities and delivered twelve training sessions. That is a modest number on paper, and it is exactly how scale should start. Two facilities are enough to prove that one module travels between sites, and that the training team can run it without outside help.
The scaling lesson is to start with two sites, not one. A single-site pilot proves the content works. A two-site pilot proves the operating model works, which is the question headquarters actually needs answered.
Fiat TOFAŞ and ErBatu Machine: Scaling Engineering Data, Not Just Training
Scale is not only about training more people. It is also about reusing the same 3D data across more teams. At Fiat TOFAŞ, RoT STUDIO implemented a perceived quality and ergonomics test system with motion capture. Design and ergonomics questions can now be evaluated in VR before physical parts exist.
ErBatu Machine took a similar route with the Designer module, moving its machinery from engineering files into immersive design reviews. RoT STUDIO's organisation profile at ITEA, the Eureka Cluster on software innovation, describes the same capability. Teams in different locations can meet in one VR scene for a design review, without travelling.
The scaling lesson is that the same model serves engineering, training and sales. Once a product exists in VR, it can be reviewed by a design team in one city, rehearsed by operators in another and shown to customers anywhere.
Beyond the Workforce: Municipal and Academic Scale
The furthest level of scale reaches people who are not employees at all. With Mersin Metropolitan Municipality, RoT STUDIO used mixed reality to help children experience safe behaviour before, during and after an earthquake. The aim was to build correct reflexes, which reading and lectures rarely achieve. The scenarios run on RoT STUDIO's Earthquake Awareness Training.
In medical education, Gazi University Faculty of Medicine carried its anatomy lessons into VR with the Designer module. The case study describes it as the start of the first local VR content development platform. It shows the same principle as Craiova: the institution builds and owns its content.
These programmes matter to corporate buyers for one reason. They prove that the same platform holds up with audiences who have never used a headset. That is the situation at every new site in an enterprise rollout.
The Scaling Playbook Behind These Programmes
Across steel, lime, automotive, public services and medicine, the programmes that scaled share a small set of decisions. None of them depends on a particular industry.
Each step below appears in at least one of the cases above.
- Start where the bottleneck is costliest: Craiova started with assembly, where one vehicle limited how many people could train.
- Build from engineering data: CAD and 3D models already exist in most plants. Reusing them makes the second module faster than the first.
- Decide who owns the content: plant-owned authoring through the RoT STUDIO License suits frequent change, while ready-made modules suit common hazards.
- Centralise management, localise scenarios: Carmeuse kept one platform and one record while each factory trained in its own context.
- Size the fleet with the three questions: locations, users and deadline decide the headset count, as PwC's deployment showed.
- Make the first session comfortable: ready-made modules such as HSE Risk Hunt Training use teleport navigation, and new users are guided to rely on it.
- Measure decisions, not completions: record errors, hazards found and choices made, then compare sites against each other.
The sequence matters less than the fourth decision. Programmes that skip central ownership tend to rebuild the same content site by site, which is exactly the drift that prevents scale.
What Scale Does to the Economics
VR content costs more to build than a classroom course. PwC found its custom VR course cost 47 percent more to create than the classroom version. The picture changes with the number of learners, because VR training is faster and needs no travel.
In the PwC study, VR reached cost parity with classroom training at 375 learners. At 3,000 learners it was 52 percent cheaper than the classroom, and at 10,000 learners 64 percent cheaper. Against e-learning, parity arrived at 1,950 learners. For a multi-site employer, those thresholds are often passed within a single programme year.
That is why the ownership question matters so much. A module built once and run across every site spreads its cost over the whole workforce. A module rebuilt for each site never reaches the learner numbers where VR pays back.
Every scaled programme above began with a clear first module and a plan for who would own it. Talk to RoT STUDIO about your sites, and we will map which delivery model fits each one.
The right model is rarely the same for every location, which is the decision the next section addresses.
Choosing a Delivery Model for Each Site
Enterprises rarely use one delivery model everywhere. The cases above combine three, chosen by how often the content changes and how specific it is to one site.
| Delivery model | Best fit | Seen in |
|---|---|---|
| RoT STUDIO License (Trainer and Designer) | Frequent change, plant-owned content, engineering data | Ford Otosan Romania, Tosyalı, ErBatu, Gazi University |
| VR training catalogue | Common hazards across many sites, fast rollout | VR Risk Hunt at the Ford Otosan event in Türkiye |
| Customized VR/XR services | Site-specific environments, centrally managed programmes | Carmeuse Türkiye, Fiat TOFAŞ |
The ready-made modules sit in the HSE training catalogue, and the trade-offs between building and buying are covered in our guide to custom VR training levels. For platform selection criteria, see choosing a VR training platform.
RoT STUDIO works from its European headquarters in Eindhoven, the Netherlands, with offices in Istanbul and Ankara, Türkiye. That footprint is built for the cross-border programmes this guide describes.
Frequently Asked Questions
These are the questions L&D, HSE and operations leaders ask most often when a VR pilot starts to turn into a multi-site programme.
How do you scale VR training across multiple sites?
Start with one module that solves a costly problem, prove it at two sites, then centralise management before adding more. Keep one platform and one training record, and let each site run scenarios that reflect its own equipment.
The cases above follow that order. Tosyalı began with two facilities, and Carmeuse Türkiye ran six factories through one centrally managed programme.
How many VR headsets does an enterprise programme need?
It depends on the number of locations, the number of learners and the deadline. PwC sized its first deployment at six headsets per location across 12 locations, then grew to 250 headsets across 25 locations as more content came online.
Should each site build its own VR content?
Each site should be able to adapt content, but not rebuild it from scratch. Plant-owned authoring, as at Ford Otosan Romania, works best when one central team sets the standard and local teams edit within it.
When does VR training become cheaper than classroom training?
In PwC's study, VR matched classroom costs at 375 learners and became 52 percent cheaper at 3,000. Multi-site employers usually pass those numbers quickly, provided content is reused rather than rebuilt per site.
Can VR training work across countries and languages?
Yes, provided the programme separates the scenario from the language layer. Ford Otosan's activity in Romania and Türkiye shows one group using VR in two countries for two different purposes at the same time.
What should headquarters measure across sites?
Measure decisions and errors, not just completions. Hazards found, steps done in the wrong order and time to competence can all be compared site by site. That turns VR into a management tool as well as a training tool.
From Pilot to Programme
The organisations in this guide did not scale by buying more headsets. They scaled by deciding who owns the content and building it from data they already had. Then they managed it from one place while letting each site train in its own reality.
That pattern holds whether the next site is a second factory, a plant in another country or an audience outside the company. If your pilot is ready for its second site, book a session with the RoT STUDIO team and bring your site list.
