If you’ve been on the London Underground recently, you might have noticed something a bit unusual: maintenance crews occasionally wearing a pair of smart glasses rather than consulting a laminated procedure card. Quietly, across UK and European rail networks, spatial computing is finding its first real home in public transport — not the passenger-facing, consumer-grade experience that was promised years ago, but the unglamorous, genuinely useful operational stuff.
The headline applications aren’t the ones anyone was predicting. Here’s where it’s actually landed.
Maintenance and Engineering
Track inspectors and station engineers are the clearest early adopters. The job involves following precise procedures in complex, often poorly-lit environments, with the constant risk of skipping a step or misidentifying a component. AR overlays solve a real problem here: instead of referencing a physical manual, the engineer sees the step-by-step procedure superimposed on the equipment they’re working on.
Network Rail piloted this with HoloLens 2 headsets on signal maintenance in 2024, and the results were good enough to expand the programme. The failure rate on multi-step procedures dropped, and new technicians reached acceptable competency significantly faster than with traditional training alone. The ROI case writes itself when you consider that a single signalling failure on a busy commute route costs hundreds of thousands of pounds in delays and passenger compensation.
The more recent shift is towards smart glasses rather than full headsets for routine inspection work — lighter, less obtrusive, acceptable for longer wear during an eight-hour shift. RealWear Navigator and Vuzix M400 have been the most commonly deployed for this use case, with voice-command interaction keeping hands free.
Control Room and Dispatch
In busy control rooms, spatial computing offers something different: an infinite, repositionable display surface. Staff managing complex train movements, CCTV feeds, and incident communications are limited by physical monitor count. Apple Vision Pro’s passthrough mode has turned up in a handful of operations centres as an experimental additional display layer, letting dispatchers float supplementary windows — train positioning maps, passenger count dashboards — above their physical screens.
The friction here is real. Control room work is safety-critical, and any interface change requires extensive validation before it’s considered operational. The productivity improvements are convincing on paper; the certification path is not trivial.
Passenger Wayfinding: The Harder Problem
AR wayfinding for passengers is the application everyone talks about and the one that’s made the least progress in production deployment. The reasons are familiar if you’ve followed XR for a while: you can’t mandate that every passenger owns a compatible device, the hardware burden is on the consumer, and retrofitting the necessary indoor positioning infrastructure across a large station is expensive.
What has shipped is more modest but genuinely useful. Several major European airports and rail terminals — including Frankfurt, Amsterdam Centraal, and Birmingham New Street — have deployed spatial wayfinding through regular mobile camera AR apps. Point your phone at a departure board or corridor junction, and an overlay guides you to your platform, exit, or accessible route.
TfL’s Elizabeth line stations were designed with this in mind; indoor positioning accuracy at Paddington and Canary Wharf is good enough to support reliable AR overlays without constant recalibration. The Elizabeth line wayfinding feature in Google Maps’ AR walking mode is a real, usable thing in 2026, though few passengers know it exists.
For accessibility, this is genuinely meaningful. Visually impaired passengers using spatial audio cues through their phone, passengers who can’t read English, and passengers unfamiliar with the network all benefit from a system that meets them where they are — on their own device, without requiring specialist hardware.
Platform Screen Doors and Safety Monitoring
Mixed reality is also showing up in safety systems, though in ways invisible to passengers. Computer vision systems overlaid on CCTV feeds can flag a passenger too close to the platform edge before a train arrives. This isn’t AR in the conventional sense, but it’s spatial computing doing real-time safety-critical work. Several metro operators, including Singapore’s SMRT and Seoul Metro, run production systems along these lines.
The UK regulatory environment for anything touching passenger safety is understandably conservative. Proof-of-concept systems exist at test facilities; full-scale production deployment in UK stations would require extensive ORR sign-off.
What’s Realistically Next
The honest trajectory for the next two to three years in the UK context:
Maintenance and engineering AR will continue to expand on rail, with the focus on complex infrastructure work where the procedure-overlay value is highest. Smart glasses will become more common in station control environments for supplementary information display. Passenger-facing AR will grow incrementally via mobile apps rather than headsets — any operator waiting for widespread consumer headset adoption to drive a wayfinding strategy is waiting for something that won’t arrive on a useful timescale.
The constraint throughout is not the technology. It’s certification, workforce adoption, and the institutional caution that any system touching safety-critical operations rightly attracts. The operators making progress are the ones who started small — one depot, one maintenance workflow, one station — and built evidence before scaling. That’s a slower trajectory than the XR industry narrative suggests, but it’s the one that’s producing real deployments.