Ancient knowledge and modern science give early warning on plant stress
An applied Indigenous-led AI case study from CSIRO - context for APS practitioners tracking sovereign AI or First Nations data governance, but not a policy item.
Key points
- Nallawilli Bunjil used CSIRO's Kick-Start program to build a machine learning vegetation classification model from drone imagery.
- The project combines Indigenous Knowledge with geospatial AI tools - an applied example of sovereign, community-led environmental monitoring.
- Limited direct relevance to APS AI governance or policy work; this is an applied research and environmental management story.
View original source
Copied.
"Ancient knowledge and modern science give early warning on plant stress"
Source: CSIRO – News
Published: (undated)
URL: https://www.csiro.au/en/news/All/Articles/2026/May/Nallawilli-Bunjil
Nallawilli Bunjil, an Indigenous-led technology company operating as BunjilView, partnered with CSIRO through the Kick-Start program to develop a machine learning model that detects and classifies vegetation from drone imagery. The 18-month collaboration produced a computer vision pipeline that converts aerial sensor data into structured land management intelligence, integrating Indigenous Knowledge of Country with geospatial analytics. A follow-on CSIRO collaboration is underway to develop hyperspectral algorithms capable of detecting early plant stress before it is visible to the human eye. The company is now expanding the model to landscapes across Australia.
Retrieved from SIMS, 18 July 2026.