4 May

Team:

The final home stretch has been a little hectic, but very rewarding for the team. We were finally able to branch out our larger power supply to a protoboard with larger 18 gauge wire, as recommended by Professor Sullivan. With our heat problems handled, we focused the rest of our time on fine tuning. Recently, we were able to get a complete solve in about 10 seconds, much faster than we originally anticipated. Lily and JT have been working on putting the final touches on the robot. Cleaning up all the wiring and housing to make the project as clean and transportable as possible. Sam has been making last minute changes to his cube state detection in order for it to run a little faster and smoother. The team decided to change the cube that will be used in the robot due to better performance in the cube state detection module and the actual robot housing. With these small changes, our team continues to perform full stack tests to ensure we have minimal errors during the final demo day. The only concern is the possibility of our robot overturning the cube. As described in JT’s section of this week’s report, we think it’s solved by swapping the current cube out for a more rigid cube.

Everyone has worked hard to get to this point. All the modules are functional and integration is nearly 100%. There were some unforeseen errors with how the pyserial module talks to the arduino through the serial monitor. But we have basic workarounds if this continues to give us problems for the demo. We seem to have most of the bases covered, and are excited to be able to finally present our cube solving robot!

JT:

This week has been all about putting the finishing touches on the project and really fine tuning the robot to the best of my abilities. This week the last of our last-minute orders came in and I was finally able to make the last minute quality of life changes I’ve been wanting to. First, Sam and I made slight adjustments to housing plans and print a new version out of perfectly clear acrylic. The housing now fits the cube a little better overall and the transparent acrylic really tops it off. On top of this, Lily and I reorganized the circuit and cable management. We sleeved all the motors so the four wires for the two coils are much more organized and compact. The general circuit layout of the project was also tidied with shorter wires. I also made the decision to change the cube we’d be using in the final demo. This is because the previous cube wasn’t rigid enough and provided problems with overturning. Lily and I originally thought the robot just wasn’t turning the desired amount of steps, but we discovered it was just the rigidity of the cube. The robot was also turning perfectly, but because the coupling arm wasn’t a perfect fit, the looseness of the cube allowed for extra slack that ultimately disrupted the turning.

The final demo is only two days away and I think we’re right on track. Our poster is complete and our project is demoable. All that’s left is to update our design document with all of our changes. With the amount of problems and minor changes we’ve had to make, we should have plenty of content to round out the rest of the final paper. Go Cubr!

Lily:

At the beginning of the week, we prepared for our final presentation (powerpoint slides) and created the design for our public demonstration poster. After the final presentation, I helped Sam with the overall software organization with respect to Python’s finicky imported module hierarchy conventions. I also assisted JT with reorganising the wiring by sleeving the motor wires to reduce the previously chaotic nature of our visual presentation. Furthermore, I branched out the connections to the larger power supply via an actual protoboard and our 18 gauge wire to improve heat dissipation.

As JT stated above, we discovered that the previous cubes that we used for physical solving were not stiff enough for our motors, which resulted in overturning at times. Luckily, we acquired several different speedcubes and were able to find one that the motors can turn as we had originally expected/desired.

I completed each step of the Beginner’s Method, but still have some bug-squashing to attend to. With respect to the cube states that are able to reach a fully solved state, the average number of moves was 84.4 moves (essentially half of our desired goal of <200 moves). I am aware of some cube states that are not able to reach a fully solved state due to bugs in the first layer algorithms, and have been going through that to quickly fix before our public demonstration this upcoming Monday.

Overall, our project is done! I’m excited for all of the careful and well-organised work we’ve done this past semester to finally come to fruition at the upcoming public demo. Go Cubr 🙂

Sam:

As this was the final stretch, we all made great efforts to add the finishing touches to our project. I validated and tested the HSV values one last time in preparation for this week. I refactored the code for better software organization and made final commits so that our team’s Github code was readable and understabdable. One last change I made to cube state detection was to hard code the scanning order so that we could scan the cube without the centerpiece caps in the Cube. This made the process of scanning and putting the robot in the housing more seamless.

I helped create the slide for the final presentation and wrote the parts of the report that were relevant to me. I helped with the remaining parts of our demo in preparation for Monday. We created the new housing using clear acrylic so that it is more visible to see the robot actually solving the cube. We also fined tuned the turn speed for the motor along with the turn delay.

Overall, we have made great progress and are ready to demo. Cubr is on track and finished on track. I am proud of all the progress the team has made and off course, go Cubr!