Modeling and simulation in reactive systems

In currently-used paradigms for system development, knowledge about the objects in the system's environment is used to derive algorithms that relate events in the environment to actions. This results in implicit representation of the original knowledge in the operational system. In the majority of cases, there are only very limited mechanisms for controlling these built-in behaviours, whereas for powerful decision support it is essential to be able to "experiment" both with various hypotheses about the behaviour of real-world entities and with the system's own possible actions. Such simulation-like operation necessitates an explicit representation of behavioural models. A new paradigm for the development of reactive systems is therefore proposed in which the behavioural models are explicitly represented and the user of the system is given direct control over these models. This allows the user to easily express both hypotheses about the intentions of real-world entities, and the system's own potential actions that are necessary to meet the mission goals. Through the control of simulated time, the possible evolutions of the environment can then be studied directly.