Camera estate triage
// What
Deploy edge vision over your existing cameras to push only meaningful events into one feed.
// Why
Your existing investment becomes useful overnight.
Change my answers
// Playbook · estate-property
Estates, property & large outdoor sites
Big sites have big perimeters and few staff. AI's natural role is sensing — knowing what's happening across your land without paying for human eyes everywhere.
// Starter projects · sequenced
Sequenced. Sized. Honest.
From a confidence-building first step to a transformative bet.
Perimeter & access-point sensing
// What
Add vehicle-detection on key entry points and animal/person classification overnight on outlying cameras.
// Why
Removes night-time alerting blind spots without hiring overnight staff.
Estate operations console
// What
All site signals (cameras + sensors + bookings + maintenance) on one console, with a dispatch workflow plumbed to whoever's on call.
// Why
Turns your estate into an instrumented system rather than a collection of cameras.
// Casework that grounds this
Built it for them.
CASE / 01
Camera-vision-enrichment — vendor-neutral scene labels
Multi-site operators with mixed-vendor camera estates
A Viseron-shaped service that watches any RTSP stream and emits structured labels (people, vehicles, animals, parcels) for downstream automation.
CASE / 02
Iris — drone detection at the edge
Ops teams running outdoor events & sensitive sites
RF + RTL-SDR + small ML on a Pi at the perimeter, alerting when a consumer drone shows up — no cloud round-trip.
CASE / 03
Halio kiosks + surface-tokens — secure dashboard delivery
Anyone needing always-on dashboards in non-secure spaces (vans, public-facing rooms)
Self-hosted kiosk hardware with HMAC-signed short-lived tokens — no shared passwords, no kiosk hijacking.
// Risks worth flagging
- Cameras pointed at footpaths, neighbours' land, public roads invite trouble — get the legal positioning right first.
- Alert fatigue will kill you faster than missed events. Tune ruthlessly.
// Watch out for in pitches
- Don't lift-and-shift a city-built model onto a rural estate; the false-positive profile is wildly different.
// The honest take
Most large-site AI projects are best framed as 'making the cameras you already paid for actually useful' before adding any new hardware.