To find the appropriate course through the buoys for the different races machine learning and neural networks will be integrated into Unity via the Unity Inference Engine, at least for the purpose of object detection from the camera feed. To simulate the boat in the water in Unity, the Dynamic Water Physics package is used. So far, we know that we will need to train our own object detection model specifically for buoys. The development approach for the actual control AI, the component responsible for generating propulsion and steering commands for the boat based on the detected objects has not yet been determined. It is possible that the team will train a neural network for this purpose, although a symbolic AI (an explicitly programmed decision logic rather than a learned model) might also be more appropriate.