High-fidelity training simulation architectures integrate all aspects of the mission system software including the system emulation, sensor/communication models, and truth engines. While this practice is beneficial for updates and maintenance, it creates difficulties when attempting to repurpose the training system to other platforms, develop new models, or reuse models with other truth engines. Additionally, managing multiple projects that require extensive integration efforts of models and truth engines requires time and money that are ill afforded on tight production schedules.

The modeling and simulation team for Boeing Surveillance and Engagement supports software test and training simulations for six intelligence, surveillance, and reconnaissance (ISR) platforms. To support its customers and requirements, the team developed a simulation architecture that separates truth, simulation models, and mission system software into different modules. These modules are tightly integrated, but loosely coupled. As a result, both mission software testing and training system configurations are reconfigurable with multiple truth engines.

This paper reviews key principles on developing a reconfigurable simulation that interfaces with multiple mission systems, expansion of sensor/subsystem models, and creating a common interface for truth data. It also provides lessons learned on how to manage the architecture to ensure future flexibility without having to create ad hoc, single-use solutions. The paper provides guidance on avoiding heavy simulation integration periods from project to project and creating simulation architectures capable of flexibility, modification, and expansion.