QCIF has partnered with the Queensland Herbarium and Biodiversity Science (QHBS) at the Queensland Department of Environment, Tourism, Science and Innovation (DETSI) through the Private Sector Pathways (PSP) program to deliver a fully operational AI workflow for wildlife camera trap image processing.

The project focused on creating a scalable, production‑ready pipeline that supports the entire process—from image annotation and dataset management to model training, large‑scale inference, visualisation and outcomes.

Built on secure, high‑performance Azure infrastructure and proven computer vision tools, the solution enables rapid, reproducible and cost‑effective species identification at scale.

It also supports the continuous improvement of models and the onboarding of new species, ensuring flexibility as monitoring needs evolve.

A collaborative pathway to modernising wildlife image identification at scale

As QCIF and DETSI worked together to develop the platform, the value of cross‑sector collaboration became clear.

“This project demonstrates what’s possible when government, research infrastructure and the private sector work closely together. By combining production‑ready AI tools with national‑scale infrastructure, we delivered a secure, scalable solution that is already transforming how wildlife camera data can be processed and used for decision‑making,” said Dr Jenna Wraith, Head of Sustainable Futures, QCIF Digital Research.

The pipeline integrates directly with national biodiversity infrastructure, including the Wildlife Observatory of Australia (WildObs), ensuring data standardisation and long‑term interoperability.

For Queensland, this capability provides a future‑proof foundation for biodiversity monitoring, conservation planning, threatened species programs and ecological research. It also offers a practical model for deploying trusted, scalable AI solutions to support environmental monitoring across Australia.

For more information about QCIF Sustainable Futures and to get in touch, visit our website.