While our design aims to deliver a fully functional and efficient autonomous bowling system, several limitations are expected due to project scope and available resources. Hardware constraints such as motor precision, camera resolution, and mechanical alignment may affect accuracy and consistency of pin targeting. Software performance is also limited by processing speed and the complexity of real-time image analysis within MATLAB. Additionally, budget and time restrictions may limit opportunities for extensive testing and refinement. Despite these challenges, our team is committed to optimizing performance within these constraints to demonstrate a reliable and innovative final system.
For project 1, there was nothing that we needed to build with outside resources since it was majority software related. The image to the right was the circuit that we set up in order for our Arduino to properly communicate with the servo motor via MATLAB.
For project 2, one issue that we had early on was the issue of securing the camera at a certain position above the paper so that it could stay stable and with enough light to pick up the shapes properly. With some common PVC from a hardware store, John was able to put together this fixture that secured the flash light making it easier for the camera to make out the shapes.
Project 3 consisted of using our vision software to find and recognize the objects on our lane and adjust our motor to the proper angle to point to the selected object. The whole process of converting the position of the object to an angle was not the biggest issue. The largest issue was that the lane was reflective so it made the object recognition very difficult. For the purpose of the demo we put a piece of white paper over the lane to properly demonstrate that the motor would be able to align with the selected object.
Going into the final design we had two main things that we needed to address. The first being converting the project 3 angular motor motion to a rack and pinion method. The second was to get the vision software for pin detection consistently detecting pins and where they are. The rack and pinion setup was not too difficult but the vision software was a big challenge. We ended up doing it based off of set pixel locations, if there is orange on that pixel or around it then it would detect a pin.