An assurance process for the exchange of operations research software models used for military simulation

Modelling and Simulation (M&S) have become an integral part of the evidence‐based approach to decision making used by Defence departments around the world. A key part of M&S is the use of computer‐based software that allows study of conflict and warfare. In order to assure confidence in the output, the Defence decision maker expects that any model or simulation is ‘fit‐for‐purpose’ and that appropriate Verification and Validation (V&V) have been conducted. We examine how best to exchange with confidence software models constructed to support military‐Operations Research (OR) studies within the context of a pragmatic OR approach that recognises the difficulties inherent in validating epistemological models of warfare that are incomplete or can only be approximately correct. The problem lies in justifying, in a rigorous and structured way, a model as being fit‐for‐purpose. The lessons from a specific case study lead us to propose that the OR and M&S communities adopt a standardised approach to the V&V of military‐OR models. We suggest that a level of assurance in models exchanged or co‐developed internationally could be readily achieved by adopting the current UK guidelines, although with some modifications, along with the acceptance of a definition for fitness‐for‐purpose cast in the context of OR studies.