Team Eleven/Journal
(→Day 5 Jan 11) |
(→Day 5 Jan 11) |
||
| Line 16: | Line 16: | ||
<div>Today was not a good day. Not enough was accomplished. We needed to have some preliminary ball recognition running today to pass checkoff 4, but we were unable to get it up and running before lab closed. Using opencv, we plan to filter for color using HSB values, then use blob detection to find the appropriate ball. We also made a cardboard model of our robot today. The model has shown that our design is weak in a few areas. For one, there is not much storage space, and the entire thing is on the small side. This is due to our attempt to design it so that it can turn in place in any position on the map. The general design is fine, but it may have to be scaled slightly differently. Some good news, base code for the main controller and state machine have been hammered out, along with our custom arduino communication. Adding states and behaviors should not be difficult.</div> | <div>Today was not a good day. Not enough was accomplished. We needed to have some preliminary ball recognition running today to pass checkoff 4, but we were unable to get it up and running before lab closed. Using opencv, we plan to filter for color using HSB values, then use blob detection to find the appropriate ball. We also made a cardboard model of our robot today. The model has shown that our design is weak in a few areas. For one, there is not much storage space, and the entire thing is on the small side. This is due to our attempt to design it so that it can turn in place in any position on the map. The general design is fine, but it may have to be scaled slightly differently. Some good news, base code for the main controller and state machine have been hammered out, along with our custom arduino communication. Adding states and behaviors should not be difficult.</div> | ||
==Day 5 Jan 11== | ==Day 5 Jan 11== | ||
| − | + | As for today, we made headway on the checkoff we missed yesterday and the one actually due today, but mainly we focused more on the work that needs to be done on our final design. Especially after yesterday's disappointments, we need to pull ourselves together. At the moment we are trying to make heads or tails on the encoder as well as other features of the bot's programming. Funny story about the encoders, the part that took the longest was my meandering throughout the buildings trying to find a printer for the pattern. Protip for freshmen: learn, love, live the MIT map. | |
We also worked on the computer vision, specifically the blob tracking. One of our computers crashed unexpectedly at one point which was quite sad, but no fear it came back to life! Moreover, we successfully passed our checkoff where the pegbot recognizes the ball and moves towards it... or in our case rams into it at full speed after contemplating it thoroughly. It was quite adorable, except for when it tried ramming into the metal leg of my chair. If this bot has a death wish it better keep it on hold until at least Monday. | We also worked on the computer vision, specifically the blob tracking. One of our computers crashed unexpectedly at one point which was quite sad, but no fear it came back to life! Moreover, we successfully passed our checkoff where the pegbot recognizes the ball and moves towards it... or in our case rams into it at full speed after contemplating it thoroughly. It was quite adorable, except for when it tried ramming into the metal leg of my chair. If this bot has a death wish it better keep it on hold until at least Monday. | ||
Revision as of 08:42, 12 January 2013
Contents |
Day 1 Jan 7
It's the first day of wonderful and beautiful IAP and with it came the kick off of Maslab. Team ll is comprised of two froshlings, Yanni and Valentina, and their upstairs neighbor and Junior member, Charles. For most of the team it was their first full day back home, and it began with a long and convoluted walk exclusively indoors to the meeting room (seeing as one of the members was recovering from a cold and didn't want to be outside). During the meeting, the competition task was explained and the merrily colored tubs were handed out.
After civilly discussing who would keep what of the materials at the end of the competition, we began work on the peg bot. We set up the arduino code to receive command codes from the computer. The command system is setup so that the computer can send a code and parameters over serial to the robot. The supported commands added today were set forward and backwards speed. The magnitude of those speeds however nearly broke the peg bot when it just barely flew off the table... note to team, do not test on table. Luckily however, it was just "nearly" and "barley" so no harm done and our check off went smoothly except for the part where the pegbot hit the wall... looks like checkoff two still needs work. Aside from that the only problems we encountered was a bit of weirdness with our motors but that could be seen to tomorrow.
By the time we finished I was starving. Deciding we'd done good, we headed off for some delicious ramen across the bridge.
Day 2 Jan 8
Deciding that yesterday's route to the meeting room was just unnecessarily loopy (not to mention impossible to follow on my own) we just braved the chilly outdoors. The lecture though a bit delayed gave us some things to think about or rather, gave the coders something to think about. Later, we finally made it back to our nice little table in the lab.
We experimented with the IR sensor, but could not get it to work reliably outside of a 3in range between 3 and 6 inches. Putting that off til tomorrow, we decided to go for bump sensors as our method of wall avoidance. A debounce class was added for digital reads. New commands were now supported by the arduino code, the ability to get limit switch info and to set the speed of the left and right motor separately. And with the temporary limit switches added to the front of the pegbot checkoff 2 was passed.
Oh, I got Edgerton training! Which is good considering I'm Mech E and will mostly be working on the construction stuff (if you can tell, I didn't write the coding bits, Charles did). Soon I will be able to begin working on the real deal robot (not that I don't love our dear little pegbot).
Day 3 Jan 9
Today we mostly worked on preparing the deliverables for the checkoff. Most importantly, we decided what design we want to go with and have a scale detailed drawing of it. We also made some progress with the vision code. Some proto code was written that we used to find balls in the sample image. The arduino code was also edited to move motor control into a class to reduce repeated code.
Day 4 Jan 10
Day 5 Jan 11
As for today, we made headway on the checkoff we missed yesterday and the one actually due today, but mainly we focused more on the work that needs to be done on our final design. Especially after yesterday's disappointments, we need to pull ourselves together. At the moment we are trying to make heads or tails on the encoder as well as other features of the bot's programming. Funny story about the encoders, the part that took the longest was my meandering throughout the buildings trying to find a printer for the pattern. Protip for freshmen: learn, love, live the MIT map.
We also worked on the computer vision, specifically the blob tracking. One of our computers crashed unexpectedly at one point which was quite sad, but no fear it came back to life! Moreover, we successfully passed our checkoff where the pegbot recognizes the ball and moves towards it... or in our case rams into it at full speed after contemplating it thoroughly. It was quite adorable, except for when it tried ramming into the metal leg of my chair. If this bot has a death wish it better keep it on hold until at least Monday.
