Agricultural sustainability is inherently multidimensional, spanning environmental, economic, and social domains. Each domain contains multiple impact metrics — from carbon footprint and ammonia emissions to welfare outcomes and antimicrobial use. These metrics often conflict, meaning no single management strategy excels across all dimensions.
This is where sustainability multicriteria optimisation becomes transformative. It provides a structured, rigorous way to evaluate livestock systems or indeed agricultural systems when multiple objectives must be balanced simultaneously, revealing the most feasible and well-balanced solutions.
Your Mission
You will develop a methodology for optimising multiple sustainability objectives within livestock systems. A system will be described by management decisions — feeding strategies, stocking density, genotype, health interventions — each influencing sustainability outcomes. Rather than collapsing these outcomes into a single score, you will treat each as a separate objective, positioning them in a multi-dimensional optimisation space.
Your work will:
- Evaluate normalisation needs across sustainability dimensions using real livestock datasets.
- Select and implement optimisation approaches — from Pareto-optimality to advanced methods such as genetic algorithms.
- Develop a hybrid AI framework combining machine learning with computational argumentation to both optimise and explain outcomes.
- Engage stakeholders to assess the clarity, usefulness, and real-world applicability of the model.
Why This Matters
Your work will help farmers, policymakers, and industry leaders understand not just what decisions lead to sustainable outcomes, but why — empowering transparent, explainable, and future-ready livestock management.
