The Technology Behind ToadTurret
AI-powered detection meets practical engineering to protect Australian wildlife.
Smart Detection
ToadTurret watches the area in front of it using a camera illuminated by white light. Every few seconds, the captured image is analysed by an AI model running directly on the device.
The model has been trained on thousands of real cane toad images from Australian properties. It identifies toads by shape, size, skin texture, and appearance, even in low light. When a toad is detected with high confidence, the system activates its spray mechanism.
We chose white light over infrared because it gives the AI much better colour and texture information, which translates directly into more accurate identification and fewer mistakes.
Safety First
Before any action is taken, the detection goes through a verification process designed to protect native wildlife and pets.
Layer 1: On-Device
The device's own AI model identifies the target. This is the first check, running locally for fast response.
Layer 2: Cloud Verification
When connected to WiFi, the image is sent to a cloud server running a second, independent AI model trained to catch things the first model might miss.
If either model detects a non-toad animal (a native frog, a pet, a bird) the system immediately pauses to let the animal move away. Without WiFi, the on-device AI still runs its own safety checks independently.
No AI is perfect. There is always a small possibility of misidentification and we are constantly working to improve accuracy. The system is designed to err on the side of caution: when in doubt, it does nothing.
Precision Targeting
Once a toad is confirmed, ToadTurret uses a servo-controlled nozzle to aim at the detected location. A controlled spray is delivered, targeted, brief, and precise.
Only the area around the detected toad is affected. The system tracks the toad's position in the frame to adjust its aim. You can fine-tune the targeting window via the app or web dashboard to suit your particular setup.
Connected Intelligence
When connected to WiFi, every detection is captured as a photo and uploaded to the ToadOps cloud. Your dashboard shows:
- Recent siting images with AI confidence levels
- Activity timeline (detections, safety pauses, system events)
- Device status and health monitoring
All data is stored securely and only accessible to the device owner. Automatic software updates keep every device running the latest AI model without any action needed on your part.
Set Up in Minutes
The ToadOps mobile app and web interface make device setup straightforward. Connect to your ToadTurret via Bluetooth from your phone or a Bluetooth-enabled computer, then optionally walk through WiFi setup for cloud features.
View your dashboard, detections, and device status from anywhere. The mobile app is available for iOS and Android, or use app.toadops.com.au.
Always Getting Smarter
The AI model improves with every deployment. Detection images from devices in the field feed into our training data, and new model versions are tested, validated, and deployed automatically.
As more ToadTurrets are deployed across Australia, the system learns from diverse environments, lighting conditions, and toad populations. More data means better accuracy, better species distinction, and fewer false positives.
Interested in ToadTurret?
Check out the store or get in touch to learn more.