Image by  Skye Anderson, PhD Student, School of the Environment, UQ

Australia is facing a biodiversity crisis. Species are declining at alarming rates, ecosystems are under growing pressure, and conservation resources remain stretched thin.

At the same time, we are collecting more wildlife data than ever before through camera traps and other emerging technologies. Yet much of this data goes unused, stored on hard drives, fragmented across institutions, and rarely translated into actionable insights.

The challenge is not a lack of data, but what we do with it.

Launching on June 1, WildObs is a national platform designed to change that. Using artificial intelligence alongside human verification, it rapidly identifies species in camera trap images and delivers high-quality data, enabling interoperable, findable data to be analysed across projects at scale.

A Joint Effort for Conservation

Developed through Australia’s National Collaborative Research Infrastructure Strategy (NCRIS), with data and compute hosted on the ARDC Nectar Research Cloud, WildObs brings together a multidisciplinary team of researchers and technical experts from across ecology, data science, citizen science, and project management. 

The project’s research lead is Dr Matthew Luskin, an ARC Senior Research Fellow and academic at The University of Queensland and it’s supported by QCIF Digital Research under the leadership of Dr Jenna Wraith, Head of Sustainable Futures Department and Principal Data Scientist. 

Adjunct Professor Sach Jayasinghe, QCIF’s CEO and project lead said, “This collaboration is aligned with QCIF’s mission of accelerating research excellence and societal impact under our Sustainable Futures portfolio, which focuses on developing cutting-edge digital research infrastructure capabilities in conservation, climate and biodiversity.”

WildObs’ collaborators include the Australian Research Data Commons (ARDC), Terrestrial Ecosystem Research Network (TERN), the Atlas of Living Australia, The University of Queensland, Bush Heritage Australia, Queensland Government Department of Environment, Tourism, Science and Innovation (DETSI), University of Tasmania, Australian Government Department of Climate Change, Energy, the Environment and Water (DCCEEW), University of Sydney and the Australian Museum, and numerous universities and NGOs including the World Wide Fund for Nature Australia (WWF), Birdlife Australia, Bush Heritage, and Australian Wildlife Conservancy, among others. 

Together, these organisations are driving the WildObs project forward, enhancing wildlife research and conservation efforts.

Powering Wildlife Data from Capture to Insight

WildObs infrastructure supports the full lifecycle of camera trap data through three integrated components.

The WildObs Image Platform enables researchers to upload and process imagery using computer vision AI models tailored to Australian environments, streamlining analysis.

This is complemented by the WildObs Database, which standardises and hosts camera trap observation data to ensure consistency and accessibility. This database has been developed by WildObs-QCIF. 

The WildObs platform hosts, and continues to incorporate, a range of computer vision models.

These include models developed by WildObs computer vision expert Dr Renuka Sharma, as well as models contributed by Google, the Australian Wildlife Conservancy, AddaxAI, and the University of Tasmania. 

WildObs currently hosts national-scale Australian models such as the WildObs National Model and the Australian Wildlife Conservancy's AWC135 model, as well as Google's global SpeciesNet model. 

The platform also hosts region- and habitat-specific models, including the WildObs QLD Wet Tropics Model and the WildObs K'gari Model. Additional models currently in the pipeline include the Tasmanian Species Recognition Model from the University of Tasmania and the Victorian Species Recognition Model developed by AddaxAI.

These systems also support the mobilisation of data to national repositories, such as the Tagged Image Repository on the Atlas of Living Australia, helping to maximise data sharing and impact.

With WildObs, users receive standardised, publication-ready detection data that can be immediately used in R packages or Jupyter Notebooks to track wildlife populations. 

As Dr Luskin says, “WildObs represents a game-changer for conservation, where timing matters and detecting problems early can mean the difference between recovery and extinction." 

Ready to explore what’s possible? 

Watch our pre-launch webinar series where you can access an introduction to WildObs, explore the WildObs database, and learn how to choose the right AI model for wildlife monitoring. 

Sign in to WildObs and subscribe to our newsletter to stay up to date on new features, platform updates, and upcoming releases.