Introduction and Motivation

The problem raised in the current paper is that of the though process behind an agent who seeks to discover firstly, the pursued agent’s strategy of movement and secondly, how to reach the agent as quickly as possible.

This exercise is meant to be a combination of graphics programming for simulating interaction between agents, goal finding algorithms and artificial intelligence. Thus the results from my work could be added alongside research in path finding and behaviour algorithms used for game or virtual reality agents, robotics and could even be a start for crowd simulation behaviour. Having an agent follow a given target is one issue, but having an agent follow an unknown target by detecting clues or by dynamically discovering the environment is a whole different story. This concept seems to be separating from the virtual world as it blends more into a natural behaviour of real people.

Apart from curiosity, other reasons for choosing this topic would be the opportunity of training a program to discover a world in pursuit of a goal, while preparing its knowledge base for other possible trajectories. Abstract or maybe futuristic uses for this topic would be in robotics, creating a detective machine that can scan an area and look for clues much more accurately, form theories based on previous research and work hand in hand with a detective. The same strategies can be used for a robotic companion who can play “hide and seek” with a child with communication problems, for example, or who can follow an elderly owner around the house (or an unknown location) when needed. An even more abstract idea would be of adopting the same theory in areas like medicine, where a nanobot with a tracking algorithm such as this could be trained to make it’s way around the human body and detect traces of infections or viral activity.