A foundation for semantic interoperability

Interoperability has emerged as critical "ility" for successful simulation development and deployment. The ever increasing complexity of the challenges to be addressed by simulation, the continuing high cost to develop simulations, and the difficulty of validating simulations all create pressures to create new simulation capabilities by composing interoperable, proven simulation assets. Interoperable simulations therefore deliver greater utility. But such interoperability is not easily provided. Certainly, technologies such as the High Level Architecture have addressed part of the interoperability challenge, specifically providing a basis for component simulations (i.e., federates) to physically connect and exchange data through those connections. This so-called technical interoperability is a necessary but hardly sufficient part of the full interoperability challenge. To achieve full interoperability, simulation compositions have to (in addition to technical interoperability) determine that the composed representations provide adequate, accurate, and consistent simulated representations that adhere to the principles of "fair fight". This demands that those composing simulations from interoperable components understand the meaning of the information exchanged, i.e., be able to define and measure the semantic interoperability present in the composed simulation. Semantic Interoperability is a technology whose goal is to make composition of simulations from existing models or simulations easy, reliable, and cost effective. In the state of the practice today, such higher order interoperability is assessed through applied engineering judgment. However such engineering judgment, no matter how well informed, is not adequate to deal with the complexity of today's complex simulations. As a result, credible, cost effective, comprehensive large scale simulations are not accessible to the engineering work force, although it remains critical to their success. Deploying semantic interoperability technology would enable an order of magnitude reduction in the time required to develop complex simulations. This paper summarizes our progress to date in developing a foundation for semantic interoperability technology, addressing the problem from five findings. First, that simulation interoperability is a specialized systems engineering problem, that is, it is a system of systems problem, which requires techniques unique to this special class of systems. Next, that simulation interoperability can fail according to a finite set of interoperability anomalies we have defined. Next, that our fidelity framework provides a basis for measuring and comparing representations to validate composed representations. Next, that specific simulation metadata can be captured and processed to directly provide for semantic interoperability. Finally, that simulation interoperability can realistically exist only with relatively strictly defined domains, meaning that product line engineering techniques such Domain Specific Graphical Languages, are the approach to exploiting the potential of simulation interoperability.