AI-powered recycling of livestock waste
Researchers are merging artificial intelligence with hydrothermal treatment to optimize how we handle livestock manure. A consortium led by scientists from China Agricultural University has used machine learning to predict how phosphorus and other nutrients partition between the hydrochar solid and the nutrient-rich liquid during processing. Published in Biochar, the work demonstrates that AI-guided control of processing conditions can boost nutrient recovery, produce higher-quality hydrochar, and reduce environmental pollution from agricultural waste.
Hydrothermal treatment: turning biowaste into valuable resources
Hydrothermal treatment uses hot, pressurized water to transform biowaste without the need for prior drying. The process yields a solid hydrochar, which can enhance soil structure and cation exchange capacity, and a liquid phase rich in dissolved nutrients. Compared with traditional methods, hydrothermal treatment is energy-efficient across a range of temperatures and is well-suited to recycling essential nutrients like phosphorus, a finite resource critical for crop production.
Machine learning models forecast phosphorus fate
To forecast how phosphorus behaves in both the hydrochar and the liquid, the team evaluated several machine learning models, including XGBoost, Decision Tree, and Random Forest. They paid particular attention to how calcium and iron ions influence phosphorus distribution during hydrothermal processing. Among the models, XGBoost stood out, delivering predictions that aligned closely with laboratory measurements and enabling researchers to map how changes in processing conditions shift phosphorus between the solid and liquid outputs.
Key findings shaping safer, more efficient recycling
The study found that reaction time and pH exert strong control over phosphorus recovery. Longer or shorter processing times and specific pH ranges can tilt phosphorus toward hydrochar or the liquid phase. Importantly, the addition of calcium and iron ions promoted phosphorus retention in hydrochar, reducing the likelihood of phosphorus leaching into water bodies. Advanced analyses showed that as treatment severity increased, the forms of phosphorus in hydrochar became more stable and uniform, which is advantageous for use as a soil amendment.
Implications for farming, environment, and policy
Integrating AI with hydrothermal waste treatment offers a practical tool for operators: by adjusting key variables such as pH, reaction time, and ion chemistry, practitioners can tailor nutrient recycling to produce higher-quality hydrochar while minimizing environmental leakage of phosphorus. This aligns with broader goals of sustainable farming and resource recovery, turning a waste stream into a valuable input for soils and reducing the ecological footprint of livestock production.
Looking ahead: global adoption and sustainability goals
Experts say this AI-assisted approach could inform policies and technologies for managing livestock waste worldwide, supporting carbon neutrality objectives and enhancing nutrient circularity. By providing data-driven guidance on processing conditions, the method helps ensure that phosphorus and other nutrients stay within agricultural use cycles, promoting safer, cleaner, and more resilient farming systems.
Journal reference: Ge, X.; Zhang, T.; Mukherjee, S. et al. Optimizing the conditions of biowastes hydrothermal treatment and predicting phosphorus fate in the hydrochar and liquid phase using machine learning. Biochar 7, 96 (2025).
