In order to solve this particular problem, the A* search algorithm was chosen as it is a heuristic search algorithms that helps in cases where the problem at hand is NP complete. The Puzzle can be solved using both a single move solution and a macro move solution which will involve a sequence of elementary single worker moves. This paper deals with the task to construct an intelligent agent capable of solving the Sokoban puzzle for multiple different field layouts. Despite the fact, the Sokoban puzzle is modelled as a computer game, it can be related to an automated planning problem for robots through the interpolation of artificial intelligence techniques. The computer game allowed players to push boxes around the maze to correctly place them in the right locations.
The Sokoban puzzle is based on a warehouse scenario consisting of boxes, walls and goal locations.
