Model-based design of integrated production systems: a review

Abstract

Pest resistance and water pollution are major issues caused by the excessive use of pesticides in intensive agriculture. The concept of integrated production system (IPS) has been thus designed to solve those issues and also to meet the need for better food quality and production. Methodologies such as agronomic diagnosis-based design, prototyping, and model-based design have been developed. Here we review the model-based design of IPS. We identify tools for the development of comprehensive models. Once comprehensive models have been developed, model-based design of IPS can be formulated as an optimization problem to be solved using different approaches. Thus, we review the choice of corresponding criteria, constraints, and mathematical formulations found in the literature. We analyze successful model-based design in transportation and drug development. We learn from these areas to overcome the inherent difficulties involved in the model-based design of IPS. We recommend the following major points: IPS model-based design should use integrative modeling platforms, process-oriented modeling, and object-oriented techniques to improve the genericity, modularity, reuse of crop models, and the data sharing. The spatial dimension of IPS has to be accounted for by making crop models spatially explicit. The design evaluation criteria have to be standardized and carefully chosen by promoting stakeholders involvement. The design of IPS should be formulated as a multiobjective optimization problem, for which non-aggregative approaches should be preferred. Finally, it is necessary to establish a true dialogue between prototyping and model-based design practitioners in order to test, and if necessary to improve, the proposed systems.