After implementing and scripting out the required actions to ensure the player movements and actions worked smoothly, I started on the AI. Using an adjusted version of the player model, I wrote an AI that would reference the players position, add a random modifier to offset it and move to that location. Upon arriving at the target position, if the player was no longer nearby the AI would select a new target position based on the above. If the player was within a certain range upon reaching its destination, the AI would strike, regardless of what direction the player was in, allowing the player to dupe around but remain cautious of each enemies facing.
Rigidbodies, mesh colliders and hit scripts were applied to weapons and written to only activate while playing the attack animation, to prevent dealing damage to targets when merely walking by them.
I was also experimenting on what environmental issues having an orthogonal camera on this angle would create. For the test I had used a castle model from a previous project of mine. It became clear very fast that any environmental models with height or substance to them would limit the player’s visibility.