A new international competition is inviting innovators to use artificial intelligence (AI) to improve pasture management. The challenge is hosted on Kaggle with a US$75,000 prize pool.
The initiative is led by CSIRO in collaboration with Meat & Livestock Australia (MLA) and Google Australia. Participants will develop AI models to estimate pasture biomass, which measures the amount of grass and edible plants available for livestock. Accurate biomass estimates help farmers make better grazing decisions, improve productivity, and support environmental sustainability.
The dataset for the competition includes pasture images linked to detailed field measurements. Images capture different seasons, locations, and pasture types. Measurements show plant height, greenness, and health. AI models can use the images alone or combine them with plant data for more precise predictions.
The competition also aims to estimate the amount of other plant species, such as clover. This reduces the need for manual sampling and gives farmers faster, more reliable information.
More than half of Australia’s land is used for grazing. Globally, grazing systems cover about a quarter of the Earth’s land surface. Improvements in pasture measurement could benefit livestock management and land health worldwide.
The competition could lead to new digital tools for pasture monitoring, helping farmers make informed decisions about grazing, pasture regrowth, and biodiversity management.