As part of a university module, I was tasked with implementing A* pathfinding for the bots in a simulation. Over the course of the project, I iterated multiple times, making several improvements.
The first enhancement was the addition of an influence map. This system used the same nodes as the pathfinding algorithm to measure control over different areas. Red team bots utilized my new pathfinding system, while Blue team bots used the simpler, pre-existing code, allowing for direct comparison. The influence map also provided measurable feedback on territorial control, though its benefit to the bots was limited. Bots tended to avoid enemy territory, but the AI navigation used for the influence map was not precise enough to provide significant advantages.
The final major improvement was the addition of networking using Winsock. I successfully replicated the simulation across multiple computers, ensuring accurate synchronization of positions, rotations, and scores. This allowed the simulation to function consistently in a distributed environment.
Stefan Yeung 