Master's thesis

Interaction Design Principles for Industrial XR

A grounded theory study on what interaction techniques actually work when the user is wearing gloves, working in noise, and can't afford to be confused by the interface.

Abstract texture background.
Published research

Most XR interaction research assumes a comfortable, controlled setting. Industrial environments don't offer that. This thesis asked interaction designers and developers what actually works — and why — when the physical context is as demanding as the task itself.

My role

Primary researcher and author

Team

Solo research, supervised by professor Markku Turunen

Timeline

2019–2021

Context

Master's programme in Human-Technology Interaction, Tampere University. Published May 2021.

Problem

What interaction techniques suit industrial XR environments best, and what makes XR solutions easy to adopt in field work?

Users
  • Maintenance technicians and field workers
  • Interaction designers and developers building industrial XR
  • Industrial companies evaluating XR adoption
Constraints
  • 6 participants — findings are not statistically generalizable
  • All Finnish — cultural context is a variable
  • COVID moved interviews to video conference, removing in-person observation
  • NDA limitations meant participants described publicized projects only
Discovery
  • Semi-structured expert interviews with 6 Finnish industrial XR professionals
  • Grounded theory methodology — theory emerges from data, not hypothesis testing
  • Supplementary literature review guided by interview findings
  • XR industry meetup observation and participant recruitment
Key insights
  • Speech is the most mature and accepted hands-free technology — familiar from personal devices, functional even in noisy environments with the right mic setup
  • Technology acceptance is the biggest barrier: wearables are intimidating, expensive, and make technicians feel self-conscious in front of clients
  • The smartphone on a holder, extended with touchless input, is still more practical for large-scale field deployment than most current HMDs
Hard metrics

No adoption, deployment, or field-performance metrics are claimed from this thesis.

Qualitative evidence

Six expert interviews and grounded theory coding produced the interaction categories, design principles, and literature-backed input-technique review.

Proxy evidence

The public thesis documents the research method, participant profile, coding logic, and resulting industrial XR interaction model.

Reflection

The grounded theory remains untested in a production deployment. The value is in clarifying why certain techniques fail and what an interaction designer needs to prioritise — not in claiming a shipped industrial product.

The problem

Industrial field maintenance is physically demanding. Technicians work in noise, confined spaces, hot and dusty environments, wearing PPE that limits their dexterity. Both hands are usually occupied. The job doesn't pause while they figure out an interface.

XR technology was entering industrial use quickly — maintenance, assembly, training, remote assistance — but most interaction research assumed office conditions. Gloves break touchscreens. Noise drowns out audio feedback. Broad mid-air gestures near machinery are a safety risk. The standard assumptions about how people interact with digital interfaces mostly don't transfer.

The thesis asked: what do the designers and developers who have actually shipped industrial XR know about this? What decisions have they made, and why?

Method

A grounded theory approach — the theory emerges from the data rather than testing a predetermined hypothesis. Six experienced Finnish industrial XR professionals were interviewed: four lead UX designers, a research director, and a software development engineer. All had substantial field experience and had shipped or closely observed industrial XR in the field.

Interview material was transcribed, translated, open coded, and progressively grouped into subcategories, concepts, and categories using constant comparison. A supplementary literature review followed, guided by the grounded theory rather than preceding it.

    0 step
  • Pre-interviews
    • Terminology review
    • XR industry meetups
    • Participant recruitment
    • 1 step
  • Interviews
    • 6 expert interviews
    • Grounded theory coding
    • Category formation
    • 2 step
  • Literature review
    • Guided by interview findings
    • Input technique research
    • Theory consolidation

Touchless interaction

Speech is the most mature hands-free technology and the easiest for users to accept. Familiar from personal devices, straightforward to implement, and functional even in noisy environments with the right microphone setup. Every participant described it as the most viable primary input channel. The bar is high though: "Good voice control is such that the user does not have to remember anything."

Mid-air gestures have real potential but come with reliability and safety concerns. Optical capture fails in poor lighting. Broad arm movements near running machinery are a hazard. Gloves break most current implementations. The threshold the interviews kept returning to: reliable as a touchscreen, or don't implement it at all. Direct manipulation gestures — pointing at an object to select it — outperform symbolic gestures that require the user to memorize a vocabulary.

Gaze is useful as a side-channel or implicit signal — directing system attention, confirming focus — but not as a primary selection mechanism. The precision demands are too high and the fatigue too significant for sustained use. Every participant agreed that multimodal input is required: no single modality covers all situations, users, or environments.

Safety and ergonomics

Unencumbered perception is the single most important factor. The interface has to stay out of the way of the user's awareness of their physical environment — a technician who is focused on a display instead of the machine they're servicing is a safety problem, not a UX problem.

PPE compatibility is non-negotiable. Gloves, helmets, hearing protection — the device has to work with all of them. That alone rules out most touchscreen-first design assumptions and pushes heavily toward voice and touchless freehand as the baseline.

Display readability matters more than immersiveness. A monocular pass-through display the technician can actually read beats a binocular HMD with a narrow FOV and poor outdoor brightness. Participants were clear: "It is very important to get the FOV wide enough." Users need to be able to choose the display that works for their job and conditions. For large-scale field deployment, the smartphone on a suitable holder, extended with touchless input, is still more practical than most current HMDs.

Worker empowerment and technology acceptance

Technicians need just-in-time information access — quick help mid-maintenance, not step-by-step guidance for tasks they already know. The interviews were emphatic about this. Industrial XR content often overcorrects toward explaining everything: "Videos of AR maintenance contain an awful amount of great looking 3D animations... 'We know how to open a screw. We do not need help with that.'" The interface should surface what the technician doesn't know, not rehearse what they do.

Reporting is a significant and underappreciated burden. One participant cited 2–4 hours of an 8-hour shift spent on documentation. "If 2-4 hours of an 8-hour workday is spent on reporting, it is a clear indicator that something has failed." XR has real potential to reduce this — in-situ documentation, automatic logging, photo capture — but only if reporting is designed as a first-class use case, not an afterthought.

Wearables are intimidating in ways that go beyond ergonomics. Technicians worry about looking unprofessional in front of clients, about damaging a €1,500 device in a harsh environment, about being visibly different from their colleagues. "Some of the technicians even think that they don't want to be seen using smartphones on the client's location as the client may think ill of them." Technology acceptance is individual and situational. It can be supported, but not mandated.

When it works, it works well. A technician with the right tool, arriving on site prepared and capable: "Maintenance technician feels like they are a hero because they can solve those problems." That's the experience industrial XR is capable of — and the bar to design toward.

The core theory

The grounded theory produced a central statement of what industrial XR interaction has to be:

A feasible industrial XR solution requires affordable and robust hands-free operation, with a freehand touchless alternative, of a readable display and an opportunity for eyes-free output — all in a usable and safe manner.

In practice: every design decision for an industrial XR interface should be stress-tested against those five constraints simultaneously. A solution that satisfies one while failing another isn't a trade-off — it's a deployment risk.

The three principles

The grounded theory produced three interaction design principles for industrial XR:

  • Design for focus. The technology is a tool for getting the job done, not a feature to learn. XR interaction should minimize cognitive burden — perception, cognition, physical effort — and stay in the background until it's needed. An interface that demands attention takes the technician away from the task it's supposed to support.
  • Readable display. Visual output is the only essential feedback channel; audio should be user-switchable. The display has to be visible and legible in the actual conditions of the work — bright light, outdoor, moving — not just in a demo room. Users should be able to choose their display solution based on their job and preferences.
  • Robust technology use. Industrial users have a low tolerance for unreliable tools. Every input technique implemented needs to be reliable enough that failure feels like the exception, not the norm. Multimodal input is required — speech as the primary hands-free channel, touchless freehand as the alternative, with PPE-compatible and one-handed operation throughout.

Interview evidence

The grounded theory structured 6 expert interviews into concepts, subcategories, and categories through constant comparison coding. The full table from the thesis:

Show interview categories

Scope and limitations

Six participants is a small sample. All were Finnish, which means the findings carry a cultural context that may not generalize to industrial XR adoption elsewhere. COVID moved interviews online, removing any in-person observation. NDA constraints meant participants could only describe work they were already comfortable discussing publicly. The grounded theory is also untested in a production deployment — its value is in structuring what experienced practitioners know, not in predicting what will ship in any specific industrial context. Read it as a design lens, not a guarantee.

Read it

Full thesis — Tampere University repository
Direct PDF

Next Project

Master's thesis on interaction design principles for industrial XR

Interaction Design Principles for Industrial XR

A research-led set of principles for XR in industrial settings — where gloves, noise, and physical load make most interface assumptions wrong.