TL;DR
Problem Statement
Training creators in industrial settings face real challenges - designing content that’s scalable, usable, and effective on AR/VR platforms.
This project explored a central question -
How can XR be designed to support hands-on training in safety-critical environments?
Conducted as a live research collaboration between the University of Nottingham and Jaguar Land Rover supervised by Dr. Kyle Harrington. I led user research & the work contributed to my Master’s thesis.
Why this research matters
Understanding the barriers to hands-on industrial training—and where XR can actually help.
Research Goals
Identify user requirements for training creators.
Develop and Test a Prototype
Evaluate Usability and Effectiveness
Recommend Improvements & Future Research
Research Process
Step 1
Literature Review
These emerged as the most critical across both academic research and industry reports -
XR tools must feel intuitive—even for non-technical users.
Training creators need flexible, accessible ways to build content.
XR tools must fit seamlessly into existing training processes, not create new silos.
Other recurring themes that shaped the design approach -
Modular content structure
Instructional clarity and feedback
Realistic pacing and user flow
Performance tracking support
Scalability and cost-efficiency
Device comfort and long-term usability
Inclusivity and accessibility
Resistance to XR adoption
Step 2
Expert Interviews
Interviewed 7 experts from academia and industry to understand real-world training workflows, pain points and expectations.
A breakdown of what emerged from interviews across both groups, and what that meant for our XR design direction.
Step 3
Consolidating User Requirements
22 URs identified - Prioritised based on relevance, feasibility, and impact — tagged by Criticality
Improved usability with user-friendly tools and frameworks
Usability for individuals with disabilities
Assistance and guidance for varying expertise levels
Ease of use and onboarding
Support for ongoing learning and troubleshooting
Tailored to individual learning styles
Compatibility with existing systems
Effective integration with conventional training techniques
Handling various content formats
Scalability
Advanced development platforms
Adoption of advanced, less intrusive XR technologies
Training content must be relevant and engaging
Design with clear objectives & assessment mechanisms
Modular and adaptable content
Analysing learning outcomes
Effective simulation of real-world hazards & safety procedures
Reliable training methods for hazardous environments
Providing real-time, actionable feedback
Monitoring user performance metrics
Tracking user interactions
Advanced interaction techniques in XR
Enhanced immersion & realism
Step 4
Prototyping
Tool Selection
Compared Unity, ZapWorks, Torch AR, and Reality Composer.
Chose Reality Composer based on -
Fast iteration cycles — ideal for rapid prototyping
Intuitive interface for real-time AR scene editing
Wide iOS compatibility (accessible across test environments)
Low learning curve — usable by non-technical stakeholders
This choice aligned with key requirements: ease of use, quick feedback loops, and support for guided learning experiences.
Prototype Goals
Design a low-fidelity AR prototype to test the most critical user requirements -
Object interaction
Guided task flow
Contextual feedback
Completion tracking
Prototype Highlights
Faster onboarding, smoother workflows, and a design system that’s ready for whatever comes next.
Drag-and-drop UI for object interaction
Enabled intuitive manipulation of virtual elements — testing interaction design early.
Task segmentation with guided AR scenes
Allowed users to follow structured, step-by-step instructions.
Visual overlays and instructional audio
Provided multimodal feedback — critical for supporting varied expertise levels.
Completion indicators
Gave clear cues on task status — helping users self-assess progress and reduce confusion.
Step 5
Testing the Prototype
Tested with 6 participants (5 academic researchers, 1 industry expert)
Scenario - Simulate an equipment shutdown using the XR training prototype
Key Outcomes
100% task completion
5/6 users completed the task without assistance
Avg. task time - 4 minutes 28 seconds
Avg. error rate - 1.8 errors per user
SUS Score - 66.25 (Marginal Usability)
NASA-TLX - 45.8 (Moderate Mental Load)
Interpreting the Scores
Task completion and independent performance were strong — users could complete the full workflow even with limited exposure.
Moderate SUS score suggests room to refine usability — particularly in onboarding and first-use clarity.
Cognitive load stayed in a manageable range, indicating the prototype wasn’t overwhelming even in technical tasks.
SUS ratings grouped by user background. Green = Non-Designer, Orange = AR Learner, Blue = Content Writer
Where the research led us
The prototype showed promise, but revealed a few usability gaps worth solving.
Tasks were broken down clearly, making steps easy to follow.
Visual guidance boosted user confidence and reduced hesitation.
Modular design enabled reuse across different training flows.
Users wanted more guidance when making mistakes.
Onboarding was not intuitive for first-time users.
Lacked real-time analytics for instructors or evaluators.
Curious about the research process or design decisions?
Feel free to reach out
