Detecting behaviours within HLA distributed simulations with added analysis components

Simulation is a powerful tool for the study of complex systems, especially in the aerospace context, where systems are too expensive and too critical to allow real experimentation. Among simulation paradigms, a particularly well-suited approach in this context is distributed simulation, where each subsystem is simulated by its own simulation component, and the interactions between these components are regulated by a standard, such as the IEEE HLA standard, which is very popular in aerospace industry. A drawback of the simulation approach is that, in an aerospace context, the systems studied are so large that their simulations are quite difficult to develop and exploit. In this paper, we address this difficulty by the use of activity recognition techniques. These techniques allow providing assistance tools for the analysis of distributed simulations, with which those issues can be adressed. This paper presents these techniques, and shows two examples of their use: one in the context of hazard detection in a simulated airport environment for the exploitation of simulation data, and the other in the context of the development of a multidisciplinary simulation aiming at assessing the environmental impact of a day of traffic around an airport.