jsparmani / block-world-problem-ai

Solution to block world problem in AI Lab using bfs, dfs, depth limited search and iterative deepening

Blocks World AI

A basic AI that solves the Blocks World Problem using searching algorithms. Built in Python 3.8.

Searching algorithms

• Depth First Search (DFS)
• Breadth First Search (BFS)
• Depth Limited Search (BFS)
• Iterative Deepening (BFS)

Running the program

The program takes 3 command line arguments:

• Search algorithm: depth/breadth/limited/itdeep
• Problem file name: Choose a .txt file from the problems folder.
• Solution file name: Optional. If it's not specified the solution will be saved in a folder called solutions.

To run it simply type in your terminal the following command:

python main.py depth ./problems/Q1.txt
python main.py breadth ./problems/Q2.txt
python main.py limited ./problems/Q3.txt
python main.py itdeep ./problems/Q4.txt

Output

The output can be seen in the solutions folder. The output is a series of steps that instruct the AI machine to reach from initial state to goal state. The output move can be one out of these three -

• (from, 'table', 'on') - Which means to move a block from table to a block which is clear
• (from, on, 'table') - Which means to move a block from top to table.
• (from, below_block, to) - Which means to move a block from one place on top of another block. The from block is above below_block

Note

There is no path possible to Q3 in our case so a file won't be generated and solutions doesn't exist will be printed in the terminal