From Maslab 2011
< Team Three(Difference between revisions)
|Line 182:||Line 182:|
<!-- Title: The final Cut -->
<!-- Title: The final Cut -->
Latest revision as of 04:49, 31 January 2011
January 3rd: Where Everything Started
- Day one. Team is ready. Team is willing to show off... but uORC is not.
- After assembling the pegbot and linking the (fancy) eeePC to our PC via the bot client, we encountered a serious problem. The uORC would not respond because of firmware.
- Good news : the same program works on a previous version of the uORC (all our work is not vain !)
- Final positive event : we tore down the webcam in order to get the essential (the printed circuit) and it still works.
January 4th: The Grand Design
- Day two. Met in the morning, discussed the game. We have a plan: watermill+waterpark slides+catapult style robots with layers, just like an onion (or a cake).
- After the group interview, we put all our sensors together and tested them. Everything works! We are on a roll.
- What's next? CADed most of the components we have already today and will finish the robot architecture tomorrow.
January 5th: The Grand Design Part II
- CADing more stuff and started vision code
January 6th: The All Seeing Eye
- Yesterday involved much work, and we were all to tired to journal.
- Video code and Cinnamon Toast Crunch abounds, and we are recognizing objects and taking names. (Of the objects)
- With the rest of tonight we'll be wrapping up the PID controller using the quad encoders -- so far seems good. Tomorrow during the day, we will combine this with the camera error and be ready to win Mock 1.
January 7th: The Puppy
- Some camera/PID communication code went slightly kaput at the last minute, so we took the less elegant approach of not using PID for Checkpoint 5. We still completed it just fine, found balls, etc -- and noted that the robot looked like a puppy chasing balls.
- Everybody is still helping out with the CAD design, which is looking more and more like a finished product. Where to put the laptop, though?
- Some people will be working on finishing the CAD and fabricating the robot, while some are working on making the pegbot able to score some goals to win pizza at Mock 1.
January 8th: Sensory Saturday
- Time to place sensors on the pegbot. Lets check all of them, two encoders, gyro, bumps, IR, breakbeam.
- Audren and Faye are working around the clock to finish the robot from their CAD models. Final check on the CAD design before sending it to the lasercutter.
- ??? What is the best way to design the ball ramp? Respects to waterpark designers.
- ??? What is the best way to design the ball scooper? Thinking strongly of using a kitchen utensil, can you guess which one?
January 9th: CADing Continues
- Unsatified with the design, more modifications have been made.
January 10th: Mock 1
- Lasercutting delayed because with the preset parameters, the 1/4" acrylic didn't get cut all the way through.
- Mock 1 run does exactly what we want it to do. Spin and aim for the closest ball. We displace one. Yes!
- Filling our bellies with our own pizza, Feta and Jalapeno, Sausage and Mushroom, coding continues.
January 11th: Lasercutter Mastered
- Central Machine Shop is out of 1/4" acrylic, so we make due with 3/8". To the reader: Buy acrylic early.
- Cut, cut, cut ...
January 12th: Snow Day
- Boston Buried, We Were Trapped
- Audren walked to lab and found it empty. Proceed to assembly the camera mount.
- Jamie is coding up a storm -- PID, wall detecting, ball finding -- all a piece of cake.
- Chances at winning Mock 2 without a properly tested robot are small, but we might be able to pull it off.
January 13th pm - 14th am: Race to Mock 2
- Coding happened all night long, and a lot got done, but there are still some screw-ups, along with the finished robot product being different from the testing platform
- After debugging some crashing into wall behavior for quite some time, it was found that a wheel was loose. Hard to avoid walls with a loose wheel; that's what I always say.
- Robot completion is nearing in the AM. For Mock 2 we will not be able to score over the yellow wall, however we plan to use our incomplete robot to score in regular goals. Retrofitting of a makeshift arm+servo mechanism is underway, and could be promising.
- A Mock 2 victory is looking very distant at the moment. However, unlike Mock 1, it is a possibility.
January 14th: Mock 2
- Whole day assembly in preparation for the mock competition. No ball slide. Arm hastily put together. No drawbridge. Minimal scoring mechanism.
- Ready, Set, Charge, ---> Ram the Wall
- Wheel determined to be too low and would get stuck on carpet bumps. Plan to raise in next CAD revision.
- Before final run, Wheel Falls Off
- Hurried off to Rental Equipment Fitting. (We are going on a team ski trip)
January 15th: Extreme Makeover - The Robot Edition
- +BallSlide, ++Drawbridge
- CADing new wheel design
January 16th: Second Sunday
- Bought an Al tube as backup for the BallSlide
- Final final CAD layout
January 17th: Tinkering
- Lasercut new wheels
- Playing with all the bump sensors, Count:8
- Attempted arm design and finished drawbridge design
January 18th - 19th am: Race to Mock 3
- Finished arm design
- Remounted bump sensors because the epoxy came off.
- All-Night Code
January 19th: Mock 3
- Found out that the bump sensors were mounted above the wall, and need to lower them.
- Give the Monsieur a moustache.
- Change strategy at the last moment to WIN!
- Food was delicious, sponsors liked our robot.
January 20th: Code code code !
- As Monsieur Robot is somehow operationnal, James could test his code on it
- MechE designers chill out
January 21th: Rush before skiing
- Code keeps on improving
- Start the design of an acrylic ramp that would be more reliable than the wooden prototype
- Left lab early to catch the bus for Sunday River.
January 22nd to 24th: On the slopes by -20 degree Fahrenheit
- The team takes some rest before the last week.
- Sunday river is awesome.
January 25th: Back in the business
- Back to code improvement
- The design of the acrylic ramp is finalized
January 26th: Fixing details
- The ramp is laser cut and assembled. It works just fine
- Breakbeam sensors are built and mounted on Monsieur Robot. Now we can catch balls more reliably.
- the two front bump sensors are replaced by new ones. They are more sturdy and look like they can handle several hours of use.
January 27th: What else?
- A small modification in the code makes a huge difference. Monsieur Robot goes slower and makes less mistakes
- The wheels became loose again. We decide to change the screws with unused ones and loctite them. we never had any problem with those again.
- We impound our Monsieur robot for the night. Tomorrow is gonna be a big day for him!
January 28th: This it!
- Final competition. We finish 2nd after losing twice against team 2 at the winner bracket final and overall final.
- We did not encountered any kind of failure during the entire competition nor changed the code in between rounds. Monsieur Robot worked perfect!
- Most importantly, Monsieur Robot won the best dressed award - the secret goal of our team from the very beginning :)