A major part of my research concerns the development of a generic modelling framework for tactical production planning.

Classical lot-sizing models generally start from a predefined mathematical structure corresponding to a particular production environment: single-level or multi-level production, single-site or multi-site organisations, capacitated or uncapacitated resources, and specific product or process structures.

The Generic Planning Model follows a different approach. The objective is to describe the industrial production system independently from a particular mathematical formulation and to use this description to construct the corresponding tactical planning problem.

The framework provides a common representation of the main entities involved in production planning, including products, product structures, production and storage entities, resources, capacities, demands, production quantities, setup decisions and inventories.

Different lot-sizing problems can therefore be obtained by instantiating the same generic representation according to the characteristics of the production system under consideration.

This approach separates three complementary levels: the description of the industrial system, the mathematical formulation of the planning problem and the optimisation method used to solve it. This separation facilitates the development of reusable models and algorithms and provides a unified framework for studying different families of lot-sizing problems.

This modelling philosophy is also the conceptual foundation of the current LotSizingDataModel software development, whose objective is to provide an extensible object-oriented representation of lot-sizing instances independently of the algorithms used to solve them.