nuk/MOVaccumCleaner

A JADE Agent Vaccum Cleaner Environment based on the MO character from the movie Wall-E . The purpose can be a assignment for my SMA class, but the idea is cool.

This Project consists of a Multi Agent Environment representing a room
	where you have a Vaccum Cleaner Robot (MO) and a Dirty Generator Robot 
	(Wall-E). This room is represented as a configurablle grid. There's also
	another agent (Auto) that acts only as a Viewer to what is happening.
	All the names of the agents were based on the characters of the movie
	Wall-E from Pixar (All Rights Reserved).

In order to execute a prepared simulation you can used the class 
	br.unb.sma.MOVaccumCleaner.spike.SpikeLaucher.
In order package the application 'mvn assembly:single'

* Vaccum Cleaner Robot:
	- Name: MO (in allusion to the Microbe Obliterator of the Movie Wall-E).
	- Class: br.unb.sma.MOVaccumCleaner.MO
	- PAGE:
		- Perceptions: 
			- Environment is a bi-dimensional Grid with configurable size. 
				So its 'location' is determined by the coordinates x,y
			- Each cell of the grid can be 'Dirty' or 'Clean'.
		- Actions: The robot is able to perform the following actions:
			- Move (right, left, up or Down): which advances him one position 
				on each direction (when not on the edges).
			- Clean: Which changes the state of the floor to 'Clean'.
			- Sense: Which tells him if the current cell is 'Dirty' or 'Clean'.
		- Goals:
			- Clean where is 'Dirty'.
			- Move to find more 'Dirty' cells.
		- Environment:
			- Accessible: Since all information about the environment is 
				directly and instantaneously provided to the agent.
			- Deterministic: Since both actions have clear output results.
				Eg: Moving changes the position and Cleaning changes the state.
			- Non-Episodic: The decision of where to go is based on which 
				direction has been taken and if a corner has been reached. 
			- Dynamic: Since WALL-E keeps changing some cells to 'Dirty' at
				random.
			- Discrete: The possible states and actions are limited.
			- Multiple Agent: Both MO and WALL-E can change the state of the 
				environment.

* Dirty Generator Robot:
	- Name: WALL-E (in allusion to the Waste Allocation Load Lifter - Earth 
		Class of the Movie Wall-E).
	- Class: br.unb.sma.MOVaccumCleaner.WallE
	- PAGE:
		- Perceptions: 
			- Environment is a bi-dimensional Grid with configurable size. 
				So its 'location' is determined by the coordinates x,y
		- Actions: The robot is able to perform the following actions:
			- Move (right, left, up or Down): which advances him one position 
				on each direction (when not on the edges).
			- Mess: Which changes the state of the floor to 'Dirty'.
		- Goals:
			- Turn a 'Clean' cell 'Dirty' (with random chances).
			- Move to find more cells to mess with ('Clean' ones).
		- Environment:
			- Accessible: Since all information about the environment is 
				directly and instantaneously provided to the agent.
			- Deterministic: Since both actions have clear output results.
				Eg: Moving changes the position and Messing changes the state.
			- Non-Episodic: The decision made regarding either to mess or not to
				mess is random not taking into account any other info. 
			- Dynamic: Since MO can clean the cells wich WALL-E messed with.
			- Discrete: The possible states and actions are limited.
			- Multiple Agent: Both WALL-E and MO can change the state of the 
				environment.

* Viewer:
	- Name : Auto (in allusion to the Auto Pilot Central Computer
		 of the Movie Wall-E).
	- Class: br.unb.sma.MOVaccumCleaner.spike.Auto
	- PAGE:
		- Perceptions: 
			- Environment is a bi-dimensional Grid with configurable size. 
			- Each cell of the grid can be 'Dirty' or 'Clean'.
			- Each cell can have a MO, a WALL-E or None.
		- Actions: The robot is able to perform the following actions:
			- Show: Shows a representation of the environment.
				- For 'Empty-Clean' cells its ' '
				- For 'Empty-Dirty' cells its '*'
				- For 'MO-Clean' cells its 'o'
				- For 'MO-Dirty' cells its 'ô'
				- For 'WALLE-Clean' cells its 'w'
				- For 'WALLE-Dirty' cells its 'ŵ'
	- Goals:
		- Observe and provide a representation of the environment.
	- Environment:
		- Accessible: Since all information about the environment is directly 
			and instantaneous provided to the agent.
		- Deterministic: Auto does not change the environment.
		- Episodic: The action is only based on the current state.
		- Dynamic: Since Auto can't change the state of the environment, from
			its point of view, the environment is changed against its will.
		- Discrete: The possible states and actions are very well limited.
		- Multiple Agent: Both WALL-E and MO can change the state of the 
			environment.