Shamal AIKAUST
01 / 07
↑ ↓ to move · F fullscreen
Ideas for KAUST

Shamal AI

A live 3D digital twin of KAUST + an AI vision layer — practical ideas for Operations & Maintenance and Industrial Security & Safety.

Ahmed Saleh · Shamal AI × KAUST Visualization Lab · 2026
What we're building now

We're building a living 3D twin of KAUST

  • In progress now — drone-scanning the campus into a navigable 3D model, on Desktop & in VR.
  • Every asset, room & system pinned with its manuals, history and live data.
  • Built for safety incidents — walk responders through the scene in 3D; locate hazards, exits and shut-offs before and during an emergency.
  • One base that both Maintenance and Safety work from.
KAUST 3D twin
A navigable 3D twin of the KAUST campus — in progress
Operations & Maintenance

Inspect from the sky — let AI flag the defects

  • Drones + cameras inspect roofs, façades, solar arrays and machinery.
  • AI flags cracks, corrosion, thermal hotspots, leaks and missing parts.
  • Reports in minutes, Arabic & English — and every finding is tracked in the twin.
  • Safer — fewer people at height or in confined spaces.
Rooftop inspection
AI-DETECTED
HVAC UNIT
HVAC UNIT
HVAC UNIT
HVAC UNIT
ANOMALY · INSPECT
GEMINI VISIONASSETS · 4FLAGGED · 1

Live Gemini detection over a demo rooftop — boxes are AI-placed on real objects, not drawn by hand.

Industrial Security & Safety

Turn every camera into a safety sensor

  • PPE & unsafe-act alerts — no helmet, no vest, work at height without a harness.
  • Intrusion, restricted-zone, crowd, fire/smoke, fallen-person — the events no one can watch 24/7.
  • Drone patrol + Remote-ID for the large perimeter fixed cameras miss.
  • A live 3D command view for response — shared with Maintenance.
CCTV safety monitoring
AI-DETECTED
VEHICLE
VEHICLE
RESTRICTED ZONE
WORKER
WORKER
WORKER
WORKER
NO HELMET
GEMINI VISIONWORKERS · 6NO HELMET · 1

Live Gemini detection on a realistic CCTV-style frame — the AI flags every worker, the vehicles, the cordoned zone, and the one worker with no hard hat.

The whole campus

The same AI reads your whole environment

Standing water (mosquito & health risk), blocked drainage, waste and overgrowth — flagged and risk-scored across the grounds, feeding both Maintenance and Safety.

Raw aerial
AI-detected
RAWAI-DETECTED
Our own AI model

Ask it to find anything — live

Our own vision model — no black box, no foreign API. Pick an image or upload your own, type what to find, and it detects it in seconds. The same engine reads your buildings, your cameras and your whole environment.

Demo aerial image
One platform · start small

One twin. One AI layer. Start with a pilot.

  • Everything on your infrastructure — Arabic & English, a KAUST-based team, data stays in the Kingdom.
  • A pilot in 2–4 weeks — one building twinned + AI on a few cameras + one VR module.
  • Then scale — full facility, drone patrol, condition dashboard → a campus-wide command view.
Try the live model — the Shamal sandbox ↗
Operations command view
One shared operations & command view — the twin + live AI in one place