Automating biosecurity sampling: using robotics and computer vision in diagnostics and surveillance

This project will build and evaluate two automated tools for biosecurity. The Entomoscope is a low-cost imaging system that rapidly photographs insects and assists with identification using machine learning. The Diagnostic Scanner automatically detects, measures, identifies, and sort specimens from trap samples using robotics and computer vision.

New Zealand’s biosecurity system depends on the early detection of new species, particularly insects that could harm agriculture, forestry, or native biodiversity.

Biosecurity surveillance and diagnostics often rely on labour-intensive manual work. Staff must inspect traps, sort specimens, and visually identify organisms, tasks that require specialist expertise and significant time. These constraints limit how many samples can be processed and how quickly potential threats can be detected.

Emerging technologies such as robotics, imaging systems, and computer vision offer opportunities to automate parts of this process, but these tools have not yet been tested in a New Zealand biosecurity context.

The project will determine whether two automated tools, the Entomoscope and Diagnostic Scanner, can improve the speed and scale of processing samples. Successful outcomes could reduce manual workload and strengthen New Zealand’s capacity to respond quickly to emerging biosecurity threats.