Team Seven/Journal
Day 0: 8 Jan 2012
Discovered the wiki.
Day 1: 9 Jan 2012
Jacob worked on the Kinect all day. David set up the EEE PC and wrote some code. Josh and Logan hacked the Arduino and made some Solidworks models. We all worked on the pegbot. Martin was not yet back. Checkoff 1 was completed later in the day. Deans connectors were installed to protect the battery and make battery swaps easy. We implemented a modular motor connector.
Day 2: 10 Jan 2012
Jacob started working on color recognition for the Kinect. Logan's jellyfish were delivered. David finished removing the GUI from the EEE PC. Josh completed checkoff 2 using a short-range IR sensor. Martin worked in Solidworks all day.
Day 3: 11 Jan 2012
We established 500k baud serial communication between the Arduino and the laptop. Logan worked on motion control code. David and Jacob upgraded to OpenCV 2.3 and continued to work on the Kinect software. We built a filtered power supply to run the Kinect from the robot battery. Logan's jellyfish turned itself inside-out.
Day 4: 12 Jan 2012
Logan's jellyfish is still inside-out but the water is turning clearer. David and Jacob worked on flood fill blob detection and got it running at a respectable frame rate. Martin continued to be an awesome mechanical engineer. Josh rewired the Arduino shield and wrote a quadrature decoding interrupt handler. Logan worked on motion control code on the Arduino. We programmed spazzbot to follow balls. Unfortunately, spazzbot ejected the Kinect and laptop onto the field in its excitement at seeing a ball,
Day 5: 13 Jan 2012
David and Jacob worked on color detection and blob recognition. Josh and Logan rebuilt spazzbot to be less spazzy. We successfully followed a ball around the field. Martin finished most of the mechanical design.
Day 6: 14 Jan 2012
Assembly begins. We cut wheels and brackets on the waterjet, and tracked down a working laser cutter to make acrylic pieces. Martin is the man. We demolish a box of 12 ice cream sandwiches in just under 5 minutes. We acquire rubber bands.
Day 7: 15 Jan 2012
Vision code continues to be optimized. Blob detector hits 70fps with downsampling. First iteration frame is constructed.
Day 8: 16 Jan 2012
We further optimized vision code and added smarter wall detection. Image processing easily runs faster than the Kinect can capture images (30fps). Laser cut some new parts. Second iteration frame is constructed.
Day 9: 17 Jan 2012
Our robot sucks balls. Just not at the mock competition, since the right angle gearboxes on our drive motors assploded.
Day 10: 18 Jan 2012
Robot is redesigned to fix all the things that broke at the mock competition. We regretfully abandon our awesome Swiss motors in favor of stock drive motors since the stock motors have a more robust output shaft. Through the fires and the flames, josh drove a nail through the right-angle gearbox on our swiss-chocolate motors. This fixed the broken state of every motor we had, and we are sticking with the fancy motors! David grows annoyed by out-of-date packages in Ubuntu 11.04 and, in a fit of exasperation, reformats the laptop and installs Arch Linux.
Days 11-14: 19-22 Jan 2012
We build build build and code code code. Robot takes shape.
Day 15: 23 Jan 2012
David bricks the Arduino, but the heroic efforts of Logan and Josh restore it to working order. The robot is obstinate and doesn't drive when it's told. Problems are fixed and we do well in the mock competition, placing second. Our weak point is field navigation. A second round of evil plans commences.
Day 16: 25 Jan 2012
This time, we actually abandon our awesome Swiss motors in favor of stock drive motors. We make more CAD drawings and start rebuilding our robot for the mock tomorrow. In other areas of great changes, we redesign our state machine code, but can't tweak constants without a final robot to test.
Day 17: 27 Jan 2012
Rev IV robot is assembled. Mysterious EE problems cause the Rev IV robot to be non-functional during the third mock competition. Hardware works great, though.
Days 18-19: 28-29 Jan 2012
David and Jacob go hiking in the White Mountains and narrowly avoid getting eaten by a hibernating bear. Josh kills all the EE gremlins. Ball launcher is installed and sends balls to behind the purple line.
Day 20: 30 Jan 2012
Hardware and software testing for seeding on the 31st.
Day 21: 31 Jan 2012
We seed first with 85 points. A mysterious Arduino communication bug appears.
Day 22: 1 Feb 2012
Testing testing testing.
Day 23: 2 Feb 2012
More testing. Robot is impounded.
Day 24: 3 Feb 2012
We win first place. The Arduino bug appears in our first matches but disappears in time for the finals.
