An agent framework to explore pathfinding strategies in maze navigation problem

The planning of paths in complex, interconnected, and unknown structures, such as mazes, is a crucial topic in various fields, including artificial intelligence and robotics. Agents capable of making independent decisions require efficient navigation through mazes, and their performance can be influenced by various dynamics and features. Understanding these factors is essential not only for developing more efficient and robust navigation algorithms but also for gaining deeper insights into which attributes to prioritize in the design and implementation of autonomous agents. In this article, we propose an agent framework to analyze various navigation strategies based on the concepts of memory and visibility. Our goal is to identify the parameters that impact the agents’ performance the most and how variations on these key parameters influence agents’ efficiency on complex maze-solving.