Poacher detection
Estates suffer night-time poaching (lamping) of deer and game by armed groups with dogs, which is dangerous and hard to catch in the act.




CAM · POACHER DETECTION
STATE 1 / 4
On the camera1.1 J / hour
No alert — quiet estate field at dusk.
Nothing wakes. Only the nanopower domain and the PIR are drawing.
How the camera spots it
Motion and radar trigger at any hour, including in darkness. Wildlife is cleared locally; people are escalated, and Gen AI distinguishes an unarmed walker from figures with torches, a rifle and a dog at night, alerting on poaching.
How you would set it up
- Wake
- PIR wakes the camera; radar confirms distance and size before any capture
- Schedule
- No timed captures. Heartbeat every 24 hours. IR illumination at night.
- On the camera
- Radar drops hares, foxes and deer by size. Edge AI drops frames with no person.
- Your criterion
Report any person in the field. Say whether they are carrying a firearm or a lamp, and whether a dog is with them.
Written in plain English. The model reads every escalated frame against it.- Alert rule
ai.person_present AND (ai.firearm OR ai.lamp OR ai.dog) AND time between 21:00 and 05:00 · hold-off 10 min- Who is told
- SMS to the gamekeeper and the estate manager; push to the rural crime team.
- Evidence kept
- Escalated frames kept with position and heading; cleared walkers logged.
What you get
23:41●●●
Lamping in progress · East Field
Two figures at the woodland edge, one sweeping a torch across the field, one holding a rifle, lurcher dog beside them.
AcknowledgeOpen camera
A readable account of what triggered it, with the frame attached. No feed to watch, no captures to trawl through.