Finally the wireless is working properly. We separated the wireless parts from the program. I hired 3 other developers to resolve it, each working independently of each other. The original developers and one of the new ones both resolved it.
While this was being done, one of the developers also refined parts of the detection module (the part that measures the ball and wheel timings, and tracks the wheel). The improvements are the ball is more accurately detected at high speeds. This is when on video, the ball is just a blur and looks nothing like a ball. From testing it appears detection at any speed is 100% accurate. Also the setup procedure is now even easier.
From here, we just need to integrate the wireless into each component, then test. Because now the wireless is working, we can uncover what other parts need attention on the actual hardware. General bug-fixing is part of the final testing.
