CSIRO partners on $3 million AI project to improve food security
Australia wastes more than seven million tonnes of food each year-around a third of everything produced. CSIRO and the University of Leeds are building an AI tool that uses fermentation to turn that waste into high-quality protein for people and animals.
What's being built
The team is developing an AI system that recommends optimal fermentation conditions to produce microbial protein powder. The target is simple: scale upcycled protein and hit price points that compete with conventional sources.
CSIRO project lead Dr Kai Knoerzer summed it up well: "Working with our colleagues internationally, this project will combine AI, fermentation science and real case studies to support industry to turn that waste into sustainable protein at scale." University of Leeds professor Dr Nicholas Watson added, "For upcycled protein to have an impact on food security, it needs to compete on price with existing products."
Target waste streams
- Damaged or unharvested vegetable crops
- Grain byproducts such as canola meal and brewer's spent grain
- Byproducts from cheese making
How the AI could deliver value
Turning heterogeneous waste into consistent protein is an optimization problem. Expect models that trade off yield, amino acid profile, contamination risk, energy use, and cost per kilogram.
Input data will likely include substrate composition, pH, temperature, inoculum, aeration, and time-series growth curves. Techniques such as Bayesian optimization or active learning can reduce lab runs and zero in on viable process windows faster.
Engineering and integration notes
For industry use, the tool will need clean data pipelines and traceability. LIMS/SCADA integration, recipe recommendations with confidence intervals, and safety guardrails will matter as much as raw accuracy.
On the production side, think batch-to-batch variability handling, sensor validation, and standard operating procedures generated alongside the model's suggestions. If the output can export to existing process control systems, adoption gets easier.
Fermentation context
Fermentation has preserved bread, cheese, and wine for thousands of years. As Dr Knoerzer noted, yeast and other microbes can convert food waste into valuable products within a circular bioeconomy framework-and do it quickly.
Funding and timeline
The project runs for two years with a total budget of $3 million. It's supported by the Bezos Earth Fund's AI for Climate and Nature Grand Challenge, a $100 million initiative focused on climate and biodiversity outcomes.
Additional perspectives
Dr Nicholas Watson emphasized cost parity as the unlock for real-world impact: "We are excited to work with CSIRO and partners across the globe to bridge that gap, launching an AI platform to support the fermentation of agri-food waste."
Dr Amen Ra Mashariki, director of AI and data strategies at the Bezos Earth Fund, said the effort shows how AI can support environmental goals when guided by science and local knowledge. That balance-model intelligence plus domain expertise-will likely decide the speed of deployment.
Learn more
If you're building similar optimization workflows, an upskilling path can pay off quickly. Explore an AI certification for data analysis to sharpen experiment design and model validation.
Your membership also unlocks:
