We believe the most significant risk that could jeopardize the success of this project is defining exactly what our system will achieve. What we proposed at first was a refereeing system that will detect and identify most commonly violated fouls in pool. However, after feedback from the professor, we see that depending on the kinds of foul that we are going for, the project will either be too simple and redundant, or very complex and perhaps inaccurate. These risk is being managed by always having an open mind to our project, and be able to adapt to faculties’ advice and continuously defining our idea more and more. Some contingency plans for our idea are narrowing down our fouls to be detected to just one complex foul such as “double hits”. Another plan we have on the back of our minds is a car racing camera system for a toy racetrack, where a system of multiple cameras will track a car through a track and provide a live stream of the camera feed.
No huge changes were made to the existing design. One thing we did consider is removing the cue stick attachments (sensors) and allowing CV to detect stick to ball collisions. This is because it would be hard for sensors to detect these kinds of collisions after looking into it deeper. This change would mean that our computer vision component will be more heavily relied upon and we need more hands on that. Since Jae was to be working on the cue stick hardware, now Jae should be helping Bhavya with any help needed on the CV part or UI part.
Our schedule has been pushed back a bit (mostly for hardware side) since we are still refining our idea and did not want to put in orders beforehand. We hope to put in orders by next week (instead of this past week) to get our schedule back on track.
No photos to report.