Here’s the thing about augmented reality in enterprise: it’s been “the next big thing” since about 2015. And for most of that time, the promise didn’t match the reality. Headsets were too heavy, software was too clunky, connectivity was too unreliable, and most pilots quietly died after the initial excitement faded.
That’s genuinely changing. Not because any single breakthrough happened, but because the technology across the board has matured enough that the friction points no longer dominate the use case. And field service — maintenance, repair, installation, inspection — is where the ROI is becoming hard to argue with.
Why Field Service, Specifically?
Field service has a set of problems that AR is unusually well-suited to solve. Technicians work on complex equipment they may only encounter occasionally. They need to reference documentation, diagrams, and checklists without putting the job down. Remote expert availability is limited. And the cost of mistakes or extended repair time is substantial — a single hour of downtime on a manufacturing line can run into tens of thousands of pounds.
The traditional answer to these problems involves a combination of printed manuals, phone calls, and a lot of experience built up over years. The problem with experience is that it walks out the door when technicians retire or change jobs. The UK is sitting on a significant engineering skills gap right now, and companies are increasingly looking at technology to make less-experienced technicians more effective faster.
AR addresses this in a few ways. Step-by-step visual overlays guide technicians through procedures they haven’t done before. Remote expert connections let a specialist in an office see exactly what the on-site technician is seeing and annotate the view in real time. And digital work instructions stay current in a way that printed manuals never do.
The Tools That Are Actually Working
PTC Vuforia is the dominant platform for industrial AR at the moment. Their Vuforia Expert Capture product lets experienced technicians record themselves completing procedures while the system captures the spatial context — where their hands are, what they’re looking at, what steps they’re taking. That recording becomes a step-by-step AR work instruction that a less-experienced colleague can follow.
The key thing about Expert Capture is that it democratises procedure creation. You don’t need a technical author or a 3D modeller. The experienced technician’s own workflow becomes the training asset. One large utilities company in the UK reported reducing procedure documentation time by around 70% after switching to this approach.
Scope AR’s WorkLink platform takes a similar approach but with more emphasis on real-time collaboration. The remote expert mode is particularly well-implemented — the expert can annotate the technician’s view with arrows, highlights, and text, and those annotations track to physical objects as the technician moves. The difference between this and a video call is significant: instead of a disconnected overhead camera and a lot of “can you move left a bit?”, the expert has real spatial context.
Microsoft’s Remote Assist, running on HoloLens 2 or increasingly on mobile devices, remains the go-to for organisations already in the Microsoft 365 ecosystem. The Teams integration means remote experts can be pulled in from the existing communications infrastructure without anyone needing a separate account or platform.
The Headset Question
For a long time, HoloLens 2 was the only serious option for industrial AR in the UK. It’s still widely deployed, particularly in sectors like aerospace and defence where Microsoft’s compliance and security posture matters. But the market has diversified.
RealWear’s wearable devices are designed specifically for hands-free operation in industrial environments — they’re voice-controlled, designed for use with PPE, and rated for noisy environments where gesture control doesn’t work. In manufacturing, food processing, and utilities, they’ve become very common.
Zebra’s headsets are popular in logistics and warehousing where barcode scanning integration matters. And increasingly, companies are deploying AR on tablets or smartphones rather than headsets for technicians who don’t need hands-free operation and want the option to use familiar hardware.
The headset question is actually secondary to the software question. Getting the platform and the workflow right matters more than the exact device. Most of the leading platforms support multiple form factors, so you can start with tablets and add headsets for specific high-value roles later.
What the Numbers Look Like
The most credible published data comes from PTC’s customer case studies and the Aberdeen Group’s field service research. First-time fix rates — the proportion of jobs completed correctly on the first visit — improve by around 25% in AR-enabled deployments. Mean time to repair drops by 15-30% depending on equipment complexity. And new technician onboarding time is typically cut by 40-50%.
To be honest, these numbers vary considerably by sector and implementation quality. AR doesn’t fix a broken process — it amplifies what you already do. A company with poor knowledge management will get poor results from AR. A company with well-structured procedures and genuinely willing technicians will see results at the higher end of those ranges.
Where to Start
The most common mistake companies make is starting with the technology rather than the problem. Pick one specific type of job where documentation or expertise gaps are causing the most pain. A particularly complex equipment type, a procedure that junior technicians regularly get wrong, a job that always requires calling in a specialist.
Pilot with that job type first. Measure the before and after. Get honest feedback from the technicians actually using the system — their buy-in matters enormously, and the best pilots involve them in shaping the process rather than having technology dropped on them.
The ROI case becomes much easier to make when it’s grounded in a specific, measured improvement rather than general efficiency claims. From there, scaling to other job types and sites is substantially simpler.
The infrastructure requirement is also worth flagging early: reliable connectivity at the point of work matters a lot. If your facilities have patchy Wi-Fi, sort that first. A good AR platform on a bad network is still a frustrating experience.