Informed Search

The informed search implementation, with two more agents.

Video showing agent implementing Depth search and Greedy search as decision making algorithms.

Video showing reactive agent behavior.

Third Programming Activity

Minesweeper

Programming assignment 3 - Search

Here we have our Rock Climbing Simulator with our Agent now performing both an informed search and an uninformed search in order to find a path. For the informed search, we implemented an A Star algorithm, utilizing some boolean flags in order to reduce the complexity of it and make it more efficient. For the uninformed search, we implemented the Depth First search, again attempting to optimize it by removing repeated nodes, yet still being quite costly. Here is a quick video demonstrating each type of search on the Android 2.3 platform. We hope you enjoy it!

2-Team, Awesome-Informed and Uninformed Searches

Informed Search

Uninformed Search

Programming assignment 3 - Search

This video shows the informed and uninformed searches make by the snake to achieve the goal that is eat the apple that appear in the board.

You can select both searches when you click in anyone that you like.

Programming assignment 3 - Searches

These are our 2 videos:

Uninformed search


This video shows how Wall-e performs an uninformed search by applying the BFS algorithm.

Informed search

This video shows how Wall-e searches for cans in an informed way applying an herustic and a cost function.

7-Badger-Programming Assignment 2

The second programming activity was to solve the eight puzzle, and for this activity, we decided to use Java as the programming language. The elght puzzle is solved with breadth first search method and an A* method.

7-Badger-Programming Assignment 1, part 2

For the next delivery, we were asked to program a set of reactive agents with a defined goal. In our case we decided to create a maze and place 4 agents on it. Each one of the agents has to find a box, which is placed in a random place.

First Programming Activity FINAL

Minesweeper


This is the final ReactiveAgent that completes all the rules designed for each number of bombs that has the principal one around it.
0 bombs- opens all the boxes around it
1 bombs- opens 7 boxes around it by random from 1 - 7
2 bombs - opens 6 boxes around it by random from 1-6
....
....

8 bombs- opens 0 boxes around it.


Here is the video of a run of the ReactiveAgent

Reactive Agents Part 2

Here is the follow-up to our previous demo of a reactive agent. Among the changes are the ability to control speed, increase the number of agents to virtually any number, and fine tuning of the agent rule set.

September 27th, 2011

Our program consist in a game similar to SNAKE, in which we use reactive agents to control the game.

Each agent have the possibility to:

- Grow each time he eats a meatball.

- Avoid to collide with the bounds.

- Avoid to collide with his own body.

Here is the Actual flash-based movie, since we are developing the application in flash, we are able to insert the current working application right into the blog.

Actividad 1

September 13th, 2011

Our program consist in a game similar to SNAKE, in which we use reactive agents to control the game.

Each agent have the possibility to:

- Grow each time he eats a meatball.

- Avoid to collide with the bounds.

- Avoid to collide with his own body.

Here is the Actual flash-based movie, since we are developing the application in flash, we are able to insert the current working application right into the blog.

Agente Reactivo Entrega 2

Éste es nuestro video complementario sobre el agente reactivo ya con funcionalidad.

Recordemos que nuestro agente es Wall-e, su misión es limpiar el mundo y para lograrlo tiene que recoger latas y transformarlas en cubos de basura para depositarlas en un lugar apropiado.

Mostramos cómo se mueve Wall-e a través del ambiente:




Ahora ya puede sensar el ambiente para realizar las siguientes reglas:

4. Moverse aleatoriamente

3. Si hay lata Y n>mis_latas -> recoger lata

2. Si mis_latas = n -> ir a la base

1. Si mis_latas = n Y en la base -> comprimir y depositar

0. Si no_energia -> recarga

La prioridad es de abajo hacia arriba

Environment Selection and Reactive Agent Implementation

http://www.youtube.com/watch?v=Ob0PmnoCiwg

Environment for agents: Unity

A video showing a brief description of how unity works.

Reactive Agent Architecture

This short video is a primitive demonstration of what our rock climbing game will be like. Though the graphics are somewhat basic, the environment's infrastructure is solid and ready to be utilized by reactive agents. Likewise, the basic demonstration of the moving agent shows that we are capable of adding agent objects to our environment, leaving just the definition of hierarchical rule-based layers in order for us to have a fully functional agent. We hope you enjoy our video!

Environment for reactive agent - Mario environment

Video showing a simple test run of a reactive agent mock.

Team 5

Minesweeper

Programming assignment 1. Team # 1


We use a java game called Minesweeper :
Is a single-player video game. The object of the game is to clear an abstract minefield without detonating a mine.
The purpose of the agent is going to be play by itself reacting to the number of mines that exist in each selected field.
We have 8 rules and so far we only have implented 1 the fist one, that if the field has zero mines around it will open all the fields around the field.
In the first part of the video we show the game without agent, later on we show the agent reacting to the field selected and opened as the first rule tells only the one that have zero mines around it.

http://www.youtube.com/watch?v=SjKwhRyl_8w

Agente Reactivo

Éste es nuestro video sobre el agente reactivo!!!!

Utilizamos Netlogo como abiente de desarrollo ya que está orientado a la modelación de agentes.
Nuestro agente es Wall-e, su misión es limpiar el mundo y para lograrlo tiene que recoger latas y transformarlas en cubos de basura para depositarlas en un lugar apropiado.

Mostramos funcionalidades del Netlogo, así como la representación de Wall-e y su movimiento a través del ambiente.



Link: http://www.youtube.com/watch?v=1SVvbCEOM8s

Agent architecture

Eline ROBERT

Mahmoud NASER

Instrucciones del Blog

Este es el blog de Sistemas Inteligentes TC2011. El objetivo es que vayan documentando sus programas durante el curso y que vayan compartiendo con sus compañeros su trabajo.

Por favor, cuando suban una actividad, pongan en el titulo el titulo de la actividad y en la etiqueta pongan su numero de equipo y nombre del equipo. En este caso, yo estoy poniendo el titulo de "Instrucciones del blog" y la etiqueta "0-JustOne".

Si desean usar más etiquetas para clasificar mejor su contribución, por favor háganlo (yo aquí estoy usando la etiqueta "instrucciones").

Cualquier pregunta, por favor mándenme un correo electrónico, preguntenme en clase, o simplemente den de alta un post y yo lo contesto.

Gracias.