The Trap Network That Tells You When It Fires
On Kangaroo Island, more than 1,000 traps are fitted with Celium alert technology, creating a connected network that communicates directly to field teams and landholders via mobile apps . The system eliminates routine physical trap checks. Teams know instantly when a trap has been triggered, and operational efficiency has increased by more than 70 per cent while reducing unnecessary vehicle movements across the landscape . This is not a pilot. It is the operational backbone of the Dudley Peninsula Feral Cat Eradication program, which spans more than 40,000 hectares .
The Data Engine: 30,000 Images, Zero Manual Sorting
Surveillance does not stop at traps. Across the eradication zone, 4G-enabled remote cameras capture around 30,000 images each day . Those images are processed in near real time by the eVorta AI platform, which automatically identifies feral cats and priority native species with high accuracy . Instead of manually reviewing vast image libraries, staff receive rapid detections that guide immediate, targeted action. As cat populations decline, near real-time analysis becomes even more critical — it helps locate the last remaining individuals and confirm eradication through proof-of-absence monitoring .
Beyond Cats: A Landscape-Scale Intelligence Model
The same intelligence-led model is being applied to other threats. In the Northern and Yorke region, AI-assisted cellular cameras and wireless trap technology are improving fox and feral cat control. On Eyre Peninsula, solar-powered 4G remote cameras are monitoring grazing pressure on the critically endangered drooping sheoak grassy woodland, where only around 3 per cent of the former extent remains . The cameras provide continuous, low-effort monitoring across large landscapes, producing defensible data on which herbivore species are impacting revegetation sites and enabling adaptive management .
The Real-World Takeaway: Biosecurity as a Data Discipline
Feral pest control has traditionally been labour-intensive, reactive, and difficult to measure. South Australia’s approach flips that model. Trap alerts, AI detections, and spatial data are integrated through GIS platforms and customised dashboards, giving managers live analytics and decision-support tools to track progress and coordinate works across complex terrain . The 70 per cent efficiency gain is not just an operational metric. It is a signal that biosecurity at landscape scale is becoming a data discipline — and that has direct implications for how Australia protects both agricultural production and native biodiversity.
