SpotPredator: Off-Grid AI Predator Detection — The Austin Python Meetup

Edge AI with no internet and no power grid

The Austin Python Meetup's August session centres on a talk by Vincent (Jiahua Zhou) on SpotPredator — a farm predator detector built to run entirely on its own, with no WiFi, no power grid connection, and no subscription service behind it.

The system runs on a Raspberry Pi Zero 2 W with a Camera Module 3. A custom-trained YOLO11-nano model performs inference directly on the device, identifying Texas coyotes, foxes, and raptors, and an instant alert is radioed to the owner over long-range LoRa. It is a useful concrete example of the constraints that shape genuinely embedded machine learning: model size, power budget, and a communications link measured in kilobits.

Practical details

• When: Wednesday, August 12, 7:00–10:00 p.m.
• Where: Brazos Place, 800 Brazos Street — note the entrance is on Brazos Street, not 8th. The in-person session is hosted by Antler in downtown Austin.
• Cost: Free; RSVP through Meetup

This is a hybrid meeting with both in-person and virtual attendance. The Meetup listing's structured data marks it as an online event, which is a listing artefact — the description gives the street address and states plainly that it is hybrid. The Austin Python group is a general Python community; this occurrence is carried here because its programme is squarely machine learning.