Implementation of an Agent Based Model for Shortest Path Finding Using Fractal Decomposition in AI

With the advent of autonomous vehicles, the need for finding effective path search algorithms has become critical and several neural networks have been adopted for this purpose. However, the extensive amount of parameters and computations become a dominant problem in the deployment of these networks in occurrences of real time inference systems owing to the network latency induced due to cloud servers. However, deploying such networks on edge computing devices proves to be a challenging task given the absence of high computation and memory capabilities at the edge devices. This issue can be resolved with the help of lightweight AI algorithms. ABM, a distributed AI tool, is used to model complex environment with autonomous, cognitive agent behavior. Here, we aim to model a path finding algorithm utilizing the principles of fractals and fractal decomposition in the environment of an ABM for observing optimum behavior.