Immersive learning portfolio · three.js
Realistic 3D where it helps learning.
Five games and mini courses built with three.js. Realistic detail where noticing it is the learning, restraint everywhere else.
About the approach
About the approach
The principle
Each project follows one principle: realism where it helps learning, restraint where it doesn't. Realistic detail is used where noticing it is the learning — terrain relief under low sun, wear marks on an artifact, the visual cue of a spill — and everything that competes for attention during a task is simplified. Cinematics are short, captioned, skippable from the first frame, and never play during assessment. The design rationale, a realism matrix and a shot list are documented in every repo.
Built with three.js, real open data (NASA, USGS, Smithsonian CC0), and CC0 assets. Learning content is CC BY 4.0; code is MIT.
Evaluation status: Automated functional, accessibility and performance testing plus agent-based design reviews are complete. No real learners have tested this yet.
Best on a desktop or laptop browser. Phones are supported with lighter graphics.
- 5 playable projects
- Real data NASA, USGS, Smithsonian
- Text mode and full keyboard play
- No tracking and no sign-up
Projects
-
HybridApollo 15 · Hadley-Apennine
Lunar Field Expedition
Land at the real Apollo 15 site, read the terrain under changing sunlight, plan a sampling traverse and defend your findings.
- Subject
- Earth and space science, research skills
- Time
- About 50 minutes
- Text mode
- Keyboard
- Reduced motion
- Captions
- Mobile
Play Lunar Field Expedition Case study for Lunar Field Expedition on GitHub
-
GameNight gallery · Smithsonian scans
The Object Detective
Examine artifacts under raking light, corroborate with archival records, and write an honest museum label.
- Subject
- History, primary-source and research literacy
- Time
- About 35 minutes
- Text mode
- Keyboard
- Reduced motion
- Captions
- Mobile
Play The Object Detective Case study for The Object Detective on GitHub
-
Mini courseWarehouse floor · Near-miss replay
Hazard Hunt
Spot warehouse hazards in realistic 3D, replay a near-miss with pause-and-decide moments, and practice reporting.
- Subject
- Workplace safety
- Time
- About 45 minutes
- Text mode
- Keyboard
- Reduced motion
- Captions
- Mobile
Content note: Includes a replay of a fictional warehouse near-miss. Nobody is hurt and no injuries are shown.
-
Hybrid100 m meadow · Quadrat sampling
Field Ecologist: Sampling the Meadow
Design a sampling plan for a realistic meadow, then see the true population and how your design biased your estimate.
- Subject
- Biology and research methods (ecology sampling, bias and uncertainty)
- Time
- About 50 minutes
- Text mode
- Keyboard
- Reduced motion
- Captions
- Mobile
Play Field Ecologist: Sampling the Meadow Case study for Field Ecologist: Sampling the Meadow on GitHub
-
Mini courseMuir Inlet, Alaska · 1941 to today
Glacier Time Machine
Stand where a glaciologist stood in 1941, align repeat photos to real 3D terrain, and measure glacier retreat.
- Subject
- Earth science data literacy (repeat photography, measurement with uncertainty, multiple causes)
- Time
- About 50 minutes
- Text mode
- Keyboard
- Reduced motion
- Captions
Content note: Climate change topic, presented evidence-first with multiple causes
Play Glacier Time Machine Case study for Glacier Time Machine on GitHub