Two-way mirror — OLED display behind it; dark pixels are transparent, lit pixels overlay on reflection
Modified UV camera — CMOS sensor with IR/UV cut filter removed + UV-pass filter (Baader U or similar)
Stereo depth cameras — Intel RealSense or OAK-D for body tracking, distance calibration, and walk-by detection
UV-transmitting dome (fused silica or UV-grade acrylic) on top for all-angle sun collection
UV Illumination
Primary: Sunlight via dome + polished aluminum funnel — no coatings, raw aluminum reflects UV well
Secondary: 365nm UV-A LEDs mounted at top of funnel, pointing into reflective walls (never direct at user) — ONLY used for UV forecast projection feature, clearly labeled as simulated
No commercial solar tubes (Solatube etc.) — their coatings filter UV. DIY aluminum funnel is better and cheaper.
Display
OLED preferred — true black = fully transparent through two-way mirror, overlay only appears where needed
LCD backlight bleeds through and fogs the mirror effect
Compute
Raspberry Pi or Jetson Nano — handles camera feed, depth processing, overlay generation
Cellular modem (IoT data plan ~$10-15/mo) for payments, alerts, UV forecast data
Power
Solar panel + LiFePO4 battery — fully self-contained, no site power dependency
Low power budget: Pi + display + camera + cellular ≈ 20-30W peak, only during active sessions
Sleep mode between uses
Operating during peak solar hours = ideal match
Interaction
NFC pad built into booth — user taps phone, creates 3-hour “Reapply sunscreen” alarm. Zero cost per user, no materials.
Tap-to-pay terminal for $5 sessions
UWB beacon for summer pass holders — auto-activates on approach
Security
Cameras for evidence
Accelerometer on housing for vandalism alerts via cellular
Form Factor Evolution
Stage
Description
POV/MVP
Regular mirror on a cart + UV camera + laptop/Pi + small display. Battery powered. Roll to beaches, farmers markets, health fairs.