===================================================================================
Authors = Federico Massa, Adriano Fagiolini
Date = 02.10.2017
Note for devs = This is a generalized and modular version of Sim_Highway
===================================================================================
LocalMonitor is a software designed to work in a distributed multi-agent environment and its goal is to monitor the behaviour of an agent whose state follows a known hybrid dynamics (finite-state-machine where each state correspond to a different time-continuous controller). The transition among discrete states must be regulated by rules based on logical events, that depend on the interaction among states of 'neighboring' agents.
The software is designed to be reconfigurable by the user, that writes a UserConfig file to describe the environment and some functions that describe the motion protocol of the observed agent.
Requirements:
To build:
docker build -t localmonitor:latest .
To open a shell inside the docker image:
xhost +
docker run --env="DISPLAY" --env="QT_X11_NO_MITSHM=1" --volume="/tmp/.X11-unix:/tmp/.X11-unix:rw" -v /absolute/path/to/localmonitor/Output:/home/Output -it localmonitor:latest bash
Inside the shell, type:
build/Simulator/Simulator
to launch a simulation.
To configure the simulation, edit Input/Simulator/Simulator.json; to configure the observers, edit Input/Observers/<ID>/Observer.json; to change the viewer type, edit Simulator/SET_SIMULATOR_VARIABLES.in
When you are done, exit the docker shell with exit, and type
xhost -
In order to use a custom environment, you should add the appropriate methods to the Input folder. In the provided version of the Input folder, you can find some sample Controllers, Automatons, Observers, System Dynamics, Sensor Models and Viewers, but you can develop your own in a similar fashion. The results are usually stored inside the Output folder. The docker run command provided before mounts a volume in the host machine, so you should find the results in your local Output folder. To develop inside docker, it is recommended to use Visual Studio Code's DevContainers (https://code.visualstudio.com/docs/remote/containers)