Global AI challenge targets faster, smarter pasture measurement
29 October 2025
Australia's national science agency, CSIRO, in partnership with Meat & Livestock Australia (MLA) and Google Australia, has launched a global AI competition to improve how we estimate pasture biomass. The prize pool is US$75,000 and the challenge is hosted on Kaggle. The goal: more accurate, more usable estimates of how much grazeable pasture is available, and what it's made of.
Why this matters to management
Grazing systems cover about half of Australia's landmass and roughly a quarter of the Earth's land surface. Better biomass data means tighter control over stocking rates, reduced risk of overgrazing, and clearer reporting on sustainability and biodiversity. It also means fewer delays between field checks and decisions.
- Cut manual sampling costs and staff hours.
- Speed up rotational planning and feed budgeting.
- Create consistent, comparable metrics across properties.
- Strengthen evidence for environmental and supply-chain reporting.
How the challenge works
Competitors will train models on pasture images linked to detailed field measurements, such as plant height and greenness based on light reflectance. Dr Dadong Wang noted that each image is paired with measurements across seasons, locations, and species mixes, allowing models to learn from images alone or from images plus plant health indicators. Participants will predict total biomass and the proportion of other species, like clover.
This approach aims to improve accuracy and usability compared with current methods. It can reduce the need for manual cuts while delivering insights fast enough to guide grazing decisions in-season.
What industry leaders are saying
MLA's Michael Lee said an AI-powered, machine vision approach can reduce the time and cost tied to manual sampling. He added that rapid, accurate differentiation of biomass components improves a producer's ability to assess current pasture quality and anticipate quantity and quality into the future.
Google Australia's Scott Riddle highlighted the scale of community involvement: "By connecting CSIRO's scientific expertise and MLA's industry knowledge with the 26 million innovators on Kaggle, we're putting a global AI community to work on a longstanding problem."
Key details for decision-makers
- Prize pool: US$75,000
- Submissions close: 28 January 2026
- Core task: Predict pasture availability and species composition from images and field data
- Dataset: Diverse seasons, regions, and pasture mixes to support generalisable models
Explore the challenge on Kaggle and submit your entry before the deadline.
Operational next steps
- Nominate a lead in data/operations to trial model outputs alongside current assessments.
- Plan field validation to benchmark model estimates against on-ground measurements.
- Scope integrations with grazing dashboards and feed budgeting tools.
- Define KPIs: sampling hours saved, decision lead time, forecast accuracy, pasture condition outcomes.
Collaboration and support
This initiative is a collaboration between CSIRO, MLA, and Google Australia. It has been supported by FrontierSI (previously known as the Cooperative Research Centre for Spatial Information). Insights from the competition could become the backbone of future digital tools for more sustainable grazing systems in Australia and beyond.
Skill up your team
If you're building practical AI capability across management and analytics, see curated programs on Complete AI Training.
Images (descriptions)
- Infographic: Pasture imagery and in-field sampling feed models built by challenge participants; outputs include biomass composition estimates to support faster, informed grazing decisions and sustainable land management.
- Brahman and Brahman cross cattle in northern Queensland, Australia.
- Examples of pasture variability used in the dataset-from lush green grass to dry, sparse cover and dense clover-like vegetation-used to train models.
- Accurate biomass estimation helps manage grazing pressure and maintain pasture quality across diverse environments.
Your membership also unlocks:
