In my thesis I modelled an Ambient Assisted Living environment to test AAL monitoring systems. In the reality this kind of system setup is really expensive, sophisticated to design, and almost impossible to verify formally, so before putting the system to work, it has to be thoroughly tested.
In this way we can obtain the expected measurement results and with that we can tune the real system before it is setup in the customer's home.
In the simulator I implemented ordinary activities of a real person and a sensor network, which can measure them. The sensor network measures the environment and the simulated person s parameters, then if necessary the intelligent system can choose the right actions.
To describe the AAL environment I created a descriptor setup file, where we can define the whole system (i.e. the inhabitated flat, the inhabitants, the ambient environment, and the information collecting sensors). The simulator displays the measurements on time series diagrams and logs information in the text files too.
The simulator was implemented on Jason platform supplemented with Java classes. In my thesis I show the architecture of the simulator from the basic level to the level of agent programming. I show also the most involved parts of the implementations, the additional services what the simulator can do, the tests of the complete system and the further development opportunities.