Model-based systems engineering framework for maximizing farm operational efficiency with agricultural autonomy metrics

Feeding a growing global population on shrinking arable land demands farming systems that deliver more with less. Precision agriculture particularly the deployment of autonomous machinery offers a promising route to higher resource efficiency. However, unlike automotive or aerospace domains with codified Levels of Autonomy (LoA), agricultural studies remain fragmented and tool-specific, resulting in ad-hoc adoption of autonomous harvesters, irrigation controllers and other devices. Lacking a domain-specific framework that translates LoA concepts into actionable metrics, farmers have no objective basis for deciding when and where to automate, making it unclear which blend of manual and autonomous operations maximizes whole-farm efficiency and exposing the inadequacy of generic LoA scales for the diverse, unstructured conditions typical of field agriculture. To address this gap, we introduce an Model based systems engineering driven framework that adopts a system of systems approach and maps autonomy metrics onto agricultural tasks. We modelled a harvest workflow across two commercial farms in SysML, simulated four configurations from manual to autonomous and assessed Task Complexity, Environmental Difficulty and Human–Machine Interaction as metrics. The framework enabled like-for-like comparison and pinpointed a optimized LoA setup that cuts cost of energy consumption by ≈ 10% and time consumption by ≈ 13% relative to the best manual baseline, while revealing where automation yields the greatest marginal gains across the fields. These results provide a transferable template for strategic autonomy planning in agriculture and open avenues for refining individual subsystems such as path planning, perception and multi-machine coordination. • MBSE based framework maps agricultural workflows for autonomy analysis. • Autonomy metrics enable comparing manual vs autonomous scenarios across diverse operations & fields. • Context specific metrics and KPIs show when and where to automate in farming. • Optimzed Level of Autonomy scenario cuts cost of energy consumption by ≈ 10% and time use by ≈ 13% in harvest simulations compared to baseline manual scenario. • Framework aids planning and refinement of agri-automation systems.