Complexity associated with the optimisation of capability options in military operations

摘要

In the context of military operations, even if the intended actions, the geographic location, and the capabilities of the opposition are known, there still remain critical uncertainties that potentially have a major impact on the effectiveness of a given set of capability options. These uncertainties greatly complicate any a priori mathematical description. Forecasting approaches, for instance, depend on predicting, as accurately as possible, the most likely future environment in which a capability system is to perform optimally. If reality diverges from the forecast, then a found system solution may perform far from its optimum. In order to develop solutions that are robust with respect to forecast inaccuracies, capability options need to be assessed in a range of futures that take account of critical uncertainties. An approach that is based on this principle is scenario analysis. Unfortunately its applicability is restricted because there are practical limits to the number of possible futures that can be explored and because the future scenarios are static and usually predetermined at the onset of capability analysis. Inspired by scenario analysis we propose an approach that exploits complexity arising from the coupling between scenarios and capability options and thus provides valuable information on the design of effective and robust capability systems. The proposed approach is iterative and makes use of both soft and hard operations research methods, with subject matter expertise being used to define plausible responses to scenarios, assess performance of options, and make amendments to underperforming options in order to increase their effectiveness. In each scenario, uncertainty affects only a subset of the system-inherent variables and the variables that describe system-environment interactions. It is this scenario-specific reduction of variables that makes the problem mathematically tractable. Our approach which we call "adversarial scenario analysis" differs significantly from existing scenario analysis processes. It can be used in conjunction with other methods, including recent improvements to the scenario analysis process. To illustrate the approach, we undertake a tactical level scenario analysis for a logistics problem that is defined by a distribution network, expected throughputs to end users, the transport capacity available, the infrastructure at the nodes and the capacities of roads, stocks etc. The throughput capacity, i.e. the effectiveness, of the system relies on all of these variables and on the couplings between them. The system is initially in equilibrium for a given level of demand. However, different, and simpler, solutions emerge as the balance of couplings and the importance of variables change. The scenarios describe such changes in conditions. For each scenario it is possible to define measures that describe the differences between options. As with agent-based distillations, the solution is essentially qualitative and exploratory, bringing awareness of possible future difficulties and of the capabilities that are necessary if we are to deal successfully with those difficulties.