The field service application of spatial computing — headsets, glasses, and mixed reality tools used by technicians, engineers, and field workers — has been promising returns for nearly a decade. For most of that time the gap between promise and operational reality was uncomfortable. Mid-2026, that’s changed in specific application areas where the returns are now measurable and the deployments are at meaningful scale.

What’s worth understanding is which applications are actually working and which are still in the perpetually-promising category. The picture is more nuanced than either the enthusiastic vendors or the skeptical critics suggest.

What’s Working at Scale

Three application areas have produced measurable, scalable ROI in field service through mid-2026:

Remote expert assistance for complex maintenance and repair. A field technician with smart glasses streaming their view to a remote senior engineer can resolve issues that would previously have required the expert to travel to site. The economics for high-value equipment with relatively few experts are straightforward — every avoided travel day pays for the technology several times over.

Guided procedure execution for repetitive complex tasks. Field workers performing complex but repeatable procedures — installation, calibration, scheduled maintenance — benefit from heads-up step-by-step guidance and verification. The error rate reduction is measurable. The training requirement for new technicians compresses meaningfully.

Spatial inspection and documentation. Field workers capturing detailed spatial records of installation sites, infrastructure conditions, or maintenance work product create documentation that’s qualitatively better than photographs and notes. The downstream value in maintenance planning, warranty disputes, and quality assurance is real.

These three applications share some characteristics. They have clear, measurable productivity or quality benefits. They don’t require continuous use throughout a workday — the technology gets pulled out for specific tasks. The implementation overhead is manageable. And the workers using them have clear, immediate benefits that overcome the friction of new technology adoption.

What’s Still Not Working

Several spatial computing applications that have been promised for years remain stubbornly difficult to deploy at scale:

Continuous augmented overlays for ongoing field work. The vision of a technician walking through a facility with continuous AR overlays identifying assets, displaying status, and providing context has been a pitch deck staple for a decade. The practical reality is that the user experience friction — battery life, weight, field of view limitations, environmental constraints — still makes this impractical for sustained use in real working conditions.

Collaborative virtual workspaces for distributed engineering teams. The hardware and software have improved. The use cases are real. The cultural and operational changes required to make distributed VR collaboration a default workflow have not materialised at most organisations. The hybrid model — occasional VR collaboration for specific high-value moments — works. The “VR as default workspace” model doesn’t.

Consumer-facing AR maintenance instructions. The idea of homeowners or end customers using AR to perform their own basic equipment maintenance has appeal but the deployment economics don’t work. The customer base isn’t large enough, the cost of supporting the technology is high, and the actual customer adoption is thin.

The Hardware Has Matured Differently Than Expected

The hardware evolution has gone slightly differently from what most predictions in 2022-2023 suggested. The high-end enterprise headsets — the Hololens lineage and equivalents — have improved but not transformed. The consumer-bridge devices have been more mixed than the hype suggested. The major surprise has been the steady improvement of smart glasses for specific industrial use cases.

The current state of hardware is roughly:

• Enterprise headsets: usable for specific applications, expensive, limited continuous-wear time
• Smart glasses: increasingly capable for narrow use cases like remote expert assistance
• Consumer-bridge devices: useful in some applications but not the breakthrough category they were positioned as
• Hand-tracking and gesture interfaces: meaningfully better than three years ago, still not as good as marketing claims
• Eye-tracking and biometric integration: capability is there, application maturity is variable

The implication is that organisations doing serious spatial computing work in field service tend to have multiple hardware form factors deployed for different applications rather than committing to a single device strategy.

The Software Story Is Better Than the Hardware Story

The software ecosystem for industrial spatial computing has matured more than the hardware has. Authoring tools for creating guided procedures, integration platforms for connecting spatial applications to enterprise systems, and management tools for deploying and updating spatial software at scale have all improved substantially.

The category that’s seen the most progress is integration. Five years ago, deploying spatial computing in a field service operation typically required custom development to connect to the underlying work management, asset management, and equipment data systems. Now there are standard integration patterns and platform tools that handle most of the common connection requirements.

This is where the implementation cost has come down most meaningfully. The hardware costs have not fallen dramatically. The implementation costs — software development, integration, change management — have fallen substantially because more of the integration work is now pre-built rather than bespoke.

For more sophisticated deployments that go beyond standard patterns, integration with bespoke enterprise data sources, custom workflow logic, and tight coupling with existing field service management tools, organisations are still bringing in specialised partners. Team400 and similar consultancies that span both the spatial computing and the underlying enterprise systems have been doing more of this kind of work as deployments move beyond pilot scale.

The Workforce Change Side

The biggest non-technical factor in successful field service spatial computing deployments has been workforce change management. The technology works. The workers using it adopt it, in most cases, with reasonable willingness. What’s harder is the surrounding organisational change — the new workflows, the new performance measurement, the new training programs, the new support models.

The organisations that have got this right have invested in the workforce change side as seriously as the technology side. The organisations that treated spatial computing as a pure technology deployment have generally struggled to realise the projected benefits even when the technology itself worked.

This is a familiar pattern from other operational technology deployments. What’s specific to spatial computing is that the worker-facing nature of the technology means workforce change effects are immediate and visible. There’s less room for the technology to be deployed and then quietly underused than there is for back-office systems.

The Vendor Landscape Has Shaken Out

The vendor landscape in industrial spatial computing has consolidated over the past few years. Several venture-backed startups have been acquired, merged, or quietly shut down. The remaining players are typically either large enterprise software vendors with spatial computing extensions, specialised industrial spatial computing firms with proven deployments, or hardware-led players who have built capable software ecosystems around their devices.

The shake-out has been healthy. Buyers can identify viable partners more easily. The integration patterns have stabilised. The product roadmaps are more predictable. The capability claims are more honest.

The Honest Mid-2026 Position

Industrial spatial computing in field service has delivered measurable returns in specific application areas at meaningful scale. Remote expert assistance, guided procedure execution, and spatial inspection are mature applications with clear ROI. The broader vision of pervasive augmented overlay throughout field operations is still not realistic at production scale.

The hardware has improved without being transformed. The software ecosystem has matured. The integration patterns are clearer. The vendor landscape is healthier. The workforce change requirements are better understood.

For organisations considering spatial computing investments in field service, the practical advice is to identify the specific applications where the ROI is most likely, deploy against those applications first, build the workforce change muscles, and resist the temptation to deploy against the broader visions that vendor marketing emphasises. The narrow application deployments build the credibility and capability that broader future deployments will eventually require.

The technology category has crossed from promise to delivery in selected applications. The next phase will be about deepening within those applications rather than dramatic expansion into new ones.