Team B9: BreakTime

This is the final blog post of our project, containing the link to our Final Report.

Christina

This past week, I collected more photograph data for my computer vision portion so that I can improve the algorithms over spring break. We split up our design doc work into parts, and I put my seldom-used writing skills to use for my sections. On Wednesday, I borrowed a car to make a trip to Home Depot with Harry to pick up our parts. 10′ pipes are really tall! We cut the pipes down into more manageable parts and got help sanding from the kind people in the Makerspace. We were able to prototype the frame build, though we will need to eventually adjust the height for the camera and projector. Next, I plan on improving the CV accuracy for our demo table so that we can start integrating everything 👍

Harry

Last week, I spent a significant portion of our time contributing to our design doc. I was personally responsible for the overarching block diagram, my part of the design trade studies, and the project management section. I aided my teammates in editing and contributing for other parts of the doc. For my specific part of the project, we made a trip to Home Depot and collected the PVC pipes and associated parts. I cut and assembled a rough frame in the lab. I held up the camera to the estimated frame height and have plans to make the frame taller. I hope to have the frame ready for a full integration when we get back from break!

Sam

This was a short week for me – I left Pittsburgh on Wednesday to start an early Spring Break. Relating to Capstone, I spent a bit of time cleaning up and refactoring the code I’ve done so far before resuming development. Most of the time this week was spent working on the design documentation with my team. Back home, I’m sharpening my pool skills at a local pool hall — can’t beat the price at $7/2 hours. 😜

TEAM

The design documentation took up a lot of our time, collectively. We had a few (very long) meetings: one to divide up work, another to sync-up, and finally one to merge all our work together and finally submit. Overall, we are satisfied with the design document we submitted despite some ambiguity. We think it will serve as a very solid draft for the final report we submit at the end of the semester. I think we went through a lot this week, but we got through it together. Team above all. 😉

Christina

This week, I went to the UC pool tables to gather more test input for the CV component of our project. I had to climb on top of trash cans to do it — such dedication 🤭 –, but I was able to get an overhead image of the entire 8-foot pool table to serve as test input for the CV component (see below image). I also took exact measurements of where the balls were placed, so I can verify the accuracy of the CV program. At the UC pool tables, I was also lucky enough to find a small, 20″ pool table laying around — one that we were thinking of buying. I received permission to play around with it and, after playing around with it for several minutes, we concluded that it wasn’t what we wanted to buy because the balls were too light and didn’t mirror the physics of a full-sized pool table.

Harry

This week, our team made a major decision to excise the robot component out of our project — in other words, I was now out of work 😢. So, most of this week was realigning my goals and schedule to help Christina with the CV portion of our project. I also led most of the research we needed to do in order to buy our parts: camera, projector, PVC pipes for our frame. I also started prototyping a frame made out of PVC pipes — the team and I realized that the frame would not be a trivial part of our project because stability might become very important later on. I went to Hunt Library to borrow one of their projectors to see how bright the projector would be in the lighting conditions we expect. Although the projector we borrowed was out of our budget, it was still useful to see what ‘lumens’ and ‘resolution’ looked like with our own eyes.

Sam

This week, I implemented 2 main features: pockets on the pool table and drawing the ‘ghost’ ball of where the cue ball would strike if it were to strike the target ball (see attached screenshot for proof-of-concept). However, pockets on the pool table are currently implemented as just circles where pockets are supposed to be; but pockets in pool are a bit more complex, with openings and angles and corners. Drawing the ‘ghost’ ball of where the cue ball would strike required more physics utility equations — namely detecting the intersection between a line segment and a circle.

TEAM

As a team, we met to work on our design presentation coming up next Monday. We chose Harry to lead this design presentation. We had a very productive meeting this past Thursday, coming up with clear goals for our MVP and for our stretch goals (beyond MVP). Below is the whiteboard work from our think session.