Iowa State University Dining VR Training Simulation
A VR training simulation for burger preparation that reduces on-the-job errors, stress, and training time during peak lunch hours.
Built in Unity for Meta Quest, this Human Factors–driven training module simulates a realistic, high-pressure kitchen environment without disrupting real operations. Trainees practice burger preparation under lunch rush conditions, receiving real-time feedback and building confidence before ever touching a real order.
- Client Iowa State University Dining
- Role XR Developer, Human Factors Designer
- Timeline Fall 2024
- Tech Stack Unity 6, C#, Meta Quest Pro, XR Interaction Toolkit
- Status MVP Prototype Complete
About the Project
Iowa State University Dining faces a persistent challenge: high error rates and elevated stress during lunch rush hours. Student employees—often new to food service—make mistakes under pressure that result in wrong orders, food wastage, and inconsistent quality.
Traditional on-the-job training ties up experienced staff and creates mistakes that affect real customers.
What the Simulation Provides
- • Hands-on, interactive burger preparation training
- • Real-time feedback via in-scene robot assistant
- • Safe practice under high-pressure conditions
- • Optional eye-tracking for attention/stress analysis
- • Scalable template for other stations
Problem Statement
Iowa State University Dining experiences high turnover among student employees, creating a constant need for training. The burger station—one of the busiest during lunch rush—has a steep learning curve combining time pressure, multi-step preparation sequences, and food safety requirements.
Peak-Hour Errors
Wrong ingredient sequences, timing mistakes, food safety lapses → food wastage, customer complaints, increased stress.
Shadowing Limitations
Ties up experienced employees who should be serving customers. Role-playing with coworkers lacks realism.
Real Customer Impact
Learning on the job means real customers receive substandard service while trainees figure things out.
The Need
A controlled, repeatable, low-risk training environment that still feels like the real kitchen—where trainees can make mistakes, learn from them, and build confidence before their first real lunch rush.
Design Approach
User Research & Task Analysis
Conducted interviews with student manager and Clyde's kitchen manager to identify specific pain points.
Developed Hierarchical Task Analysis (HTA) breaking burger preparation into discrete, sequenced actions—directly informing the VR scenario structure.
Training Scenario Design
- • Trainee information screen (employee IDs, shift context)
- • "Trial scene" for VR control familiarization
- • Core burger-prep under simulated rush
- • Robot assistant real-time feedback
- • Post-training questionnaire
VR Simulation Development
Built in Unity 6 for Meta Quest Pro using XR Interaction Toolkit. Interactive elements: grill surface, bun toaster, ingredient containers, order ticket display. Time pressure mechanics with building order queue.
Human Factors Optimization
- • Cognitive load reduction via simplified UI
- • Error prevention with corrective prompts
- • Stress training with tuned audio/pacing
- • Eye tracking for attention analysis
Demo Reel
This prototype demo shows the full training loop from the trainee's perspective inside the virtual kitchen. Watch the interaction model in action: grabbing ingredients, assembling burgers on the prep station, managing the order queue, and receiving real-time feedback from the robot assistant.
Challenges
Technical Challenges
- • Collision detection issues with rapid movements
- • Scene transitions occasionally failing
- • Hand/eye-tracking accuracy degradation
- • Bun-toasting animation conflicts
Scope Constraints
- • Only burger-prep module shipped
- • Stress detection stayed conceptual
- • Limited participant pool
- • Additional stations = future work
UX Adjustments
- • More VR onboarding needed than expected
- • Users requested clearer guidance
- • Appetite for additional roles/stations
Lessons Learned
Human Factors in XR Pays Off
Aligning instruction pacing and feedback timing with cognitive limits made training feel usable. "Stressful but learnable" vs "frustrating" came down to small adjustments.
User-Centered Design Non-Negotiable
Feedback from student workers directly shaped UI clarity and scenario difficulty. My initial assumptions about intuitive were often wrong.
XR Debugging Is Its Own Discipline
Collision detection and hand tracking issues only surfaced under realistic movement patterns. Lab-only testing missed problems that appeared with natural gestures.
Stress Can Be Designed Safely
Controllable high-pressure simulations prepared participants for real rush conditions without real-world consequences. Genuine nervousness while remaining psychologically safe.
Final Outcome
The VR training simulation successfully delivered an MVP that validated the core hypothesis: immersive, Human Factors–driven training can reduce errors, build confidence, and provide a safer learning environment than traditional on-the-job approaches.
Key Outcomes
- ✓ Reduced on-the-job training errors and food wastage
- ✓ Improved trainee confidence before real shifts
- ✓ Reinforced food safety protocols and timing
- ✓ Established scalable pattern for future modules
Future Directions
- • Additional modules: grill station, barista, cashier
- • Deeper eye-tracking/biometrics integration
- • AI-powered adaptive coaching
- • Application to other industries