From Maslab 2011
January 3rd, 2011
First time writing the date this year, and I (Piper) got it right! W00! Anyway, our team began Maslab in an extremely sleep-deprived state, which made things very amusing. (I was incredibly giggly...) (I am journaling this sleep-deprived. Expect lots of exclamation points and smiley faces!)
Problems we ran into with our code: our "Hello, World!" statement won't print without being in an infinite loop. We're not entirely sure why, but this problem didn't carry over to our Drive class (which we used to get our second component of the checkoff - the robot drove forward for three seconds). This is a good thing, because interrupting the infinite loop didn't work, and we can't make infinity last only three seconds :). After a little debugging, our code successfully drove the robot forward for three seconds and stopped.
With the actual board, we were able to attach our wheels, motors, and castor to our base. We had trouble securing all our wires to the microboard, since the crimping didn't seem to clamp the wires down all the way. We also had to short our emergency stop (due to Maslab adjusting some of the code this year) before our robot could work. But worked it did! And we got a checkoff! And then it broke again! After replacing a fuse and soldering the insides of our battery clips, our attachments were more secure and the robot worked again :).
[x] Since one of our motors were initially wired backwards, our ground and power do not follow the standard color convention. We should rewire this.
[x] Double check to see that when our code decides the robot is moving forward, the robot is moving forward instead of backwards.
January 4th, 2011
Possible algorithm for robot: The algorithm runs as long as the timer is noted less than 3 minutes (180 seconds). The first ball seen will have its color noted and be saved in a variable called our_color. As soon as this is established, search the map for goals, the goals are determined as follows: if a yellow wall is seen, drive up to it and use the ir sensor to determine whether or not the depth of the wall varies along its length. If the wall does vary, save the location of the goal in a list called goals_loc. When the number of goals is greater than 2 (or if more than 30 seconds have elapsed), then begin to look for balls. Whenever a ball is found, look for the nearest goal and transfer the ball to that goal. Do this as long as the timer has not gone over 2 minutes. After two minutes, whenever a ball is found, if the distance between the ball and the nearest known wall to the opponents side is less than the distance to the nearest known goal, then save the distance and calculate a random number between 0 and 1 and save it to rand_n. Should rand_n > e ^ -(d_togoal-d_towall), throw the ball over the wall, else, throw the ball into the goal. Stop after 3 minutes.
Robot strategy: We decided that we are stronger on the course 6 side of the spectrum than the course 2 side of the spectrum. We're going to keep our robot design relatively simple, with a conveyor belt and accompanying pinball-machine-like doors to pull balls into the robot and drop them into a compartment. On the side, we will have a door that opens when told so that we can drop all our balls into the goal. We decided not to drop balls onto the other side in order to keep our robot simple so that we can focus on our code.
- Michael - Since Maslab is his main IAP commitment, he'll code in the evenings and possibly during the day.
- Shawn - Work until 4pm every day, code at night with Michael.
- Piper - Work on building between lecture (or late mornings when lecture isn't happening) and 7pm daily.
- Xavier - Work on building between lecture (or late mornings when lecture isn't happening) and 7pm daily.
- Piper did shop training today.
- Xavier did laser training today.
[x] TA gave us a suggestion for fixing the code problem we had on the first day. Try it out. [it works now! yay!]
January 5th, 2011
New robot strategy: We came to realize that our original robot design was problematic. Our design had a lot of failure modes. The conveyor belt roller would have to be very small so that the ball would be pulled up it rather than pulled away, and the doors pushing the ball onto the roller would have to be well synchronized. We decided to start from scratch, and came up with something we believe will work much better. So, yay! :)
We're going to have a scoop with a slanted arm leading down, away from the robot. We will scoop the ball, then rotate the arm up so that it rolls down the arm into our collection box. This requires another motor with a decent amount of torque, and will likely be our main expense. Our collection box will be at least 6" above the ground. Our computer, battery, and orc board will live under the collection box. The collection box itself will be slanted towards a drawbridge on the back, which we will use a servo to open and close. With this design, we believe that building will be easier, ball collection will be more predictable, and we will now be able to score over the wall!
We are currently working on a SolidWorks design.
Other labwork: We were able to successfully attach our new sensor to the orc board and calibrate it. (The distance is 16.93*output+1.624.) We decided that for consistency, we're going to be American and use inches. We got the robot to go up to a wall, sense it, stop, and turn. Thus earning Checkpoint 3! Yay! We started working on our image processing/camera, as well as making some cutouts to envision our robot better.
Other: Piper is no longer as sleep-deprived or as sick! Thus, the journaling is less scatterbrained. Bwaha. Shawn is somehow getting his thesis work done despite also doing Maslab! Xavier is sore from Ultimate frisbee (and has practice again tonight lol)! Michael is having trouble constantly switching between Mac OS, Ubuntu, and Windows.
Other other: Our print statement during testing was "Scully Scully Scully Scully Scully". This amused Michael greatly. Piper may or may not be obsessed with the X Files.
[x] Possibly move wheels more towards the middle of our board. [we're probably going to with a two-wheel-two-caster operation
[x] Figure out the torque needed on the new motor (weigh the balls?) [balls are super light]
January 6th, 2011
Lab: Shawn and Xavier worked on robot design while Michael worked on the Checkoff 4 imaging problem. Piper joined Michael after the H-bridge tutorial. Also soldered a header together so that we could short our ESTOP in a less painful way :).
Other: Piper went to the H-bridge tutorial. Below are her random notes.
- Current sense lines, always ground them. Both enable A and enable B should be tied to Vss. Input 1 and input 2 are tied to enable A and 3/4 are supplied to B. Orcboard output 5V needs to be connected to 5V line. Need to ground battery to ground and ground orcboard to ground. Supply voltage needs to be connected to the 12V line on battery.
- Control-c might stop the code but not the motor. Make sure to clear the ports so that the motor stops in time.
- Data sheet here: http://web.mit.edu/6.186/2011/Lectures/L298.pdf
- There was an email sent out about shutdown hooks. We should look here: : http://download.oracle.com/javase/6/docs/api/java/lang/Runtime.html#addShutdownHook(java.lang.Thread)
- Michael and Piper have been arguing about who's a better Mexican for the past half hour.
- And then a Spanish-speaking machine called Xavier's phone. He might be part of the Mexican mafia, and thus be most Mexican of all.
- Robot beta plan (Piper's way): Robot is a giant skewer. Robot skewers other team's robot until dead. Robot then skewers own balls and tosses skewer onto other side. Profit!
- Robot charlie plan (Shawn's way): robot is a giant vacuum. Robot vacuums up all balls. Robot fires all balls at the other team's robot. Profit!
- Violence is the answer.
- Robot delta plan (Shawn's other way): Build a plane. Collect balls upon takeoff. Dump on other side. Should be easy, right? :D
- We were really happy to realize that our journal was long, thus writing the final paper will likely be super easy. Then we realized only about a quarter of the information here would actually be useful to our papers >.> :D
[x] Figure out why our motors are so temperamental.
[x] Decide where we want our sensors and camera.
January 7th, 2011
Labwork: Our code can find red, blue, green, and yellow things! Yay, Checkoff Four! We also now have a full cardboard mockup of our robot, so we can start building. We're currently hard at work on the Checkoff Five code.
- CODE CODE CODE
- SolidWorks ? [decided not to do this because we have a physical model]
- Use Monday's mock competition as a place to test our detection code and robot movement
- Make robot parts and build this week!
January 10th, 2011
Lab: SO MUCH CODING!!!!!! We did a lot of coding today, as well as revised our strategies for post-mock-competition. (Our robot isn't fully built yet, so our idealized code wouldn't have worked for this mock competition.) We ran into a lot of problems getting our camera to work :( Hopefully we'll be able to spend Tuesday-Thursday building and constructing for our robot, and testing our more finalized code by Thursday evening.
- New robot strategy: Our robot will totally not be a trained puppy with an orcboard taped to it. It may look that way, but we promise the drool is purely synthetic.
- "How can Farrah Fawcett die? She was so cool!"
- Batman hasn't died because he is actually cool.
[x] Figure out why our repository is sad. (We ended up not using the repository.)
January 11th, 2011
Lab: EVEN MORE CODING OMGZ!!!! We also made some progress on making our design REAL! We have dimensions ready and hinges designed so that we can spend the rest of the week building, hoping to be done by Thursday evening. We got the camera to supply tracking! Now it accurately gives information to the controller and gives recommendations on what way the robot should turn.
Michael then doubled the speed of image processing. We were doing the main process twice. Other:
- Beakman is better than Bill Nye, but no one ever saw Beakman :/
- Xena was way more attractive than Bill Nye.
- We have been in Maslab forever.
- Piper hates computers. (But she really loves them.)
- The passage of time is truly a mystery. "How did it 6pm?!"
- Michael is obsessed by the song "Vagabond" from 500 Days of Summer (a movie that makes him cry).
January 12th, 2011
Snow day!!!! Michael did some angle from image data calculations and got stressed when he couldn't work it out so he came to annoy Piper.
January 13th, 2011
Lab: Since lab was closed yesterday, we're now working on getting our pieces for the bot today. We decided to go with plexiglass, which is a bit more difficult to cut, so we've changed to lasering it. Our robot can now hopefully work its way ok around walls. Yay for metalwork and running to the hardware store to get hinges. (If only Economy Hardware weren't closed :(.)
- Descent was the best game ever.
- Scully is really, really short. Like, 5th grader short. And she's the same height as Piper.
January 14th, 2011
Lab: After many interesting events last night/this week where the world just did NOT want us to build, we finally got our lasercutting-of-the-plexiglass done! Yay! We also made l-bars and other things to prepare for finally building this thing, which we will do tomorrow. We got checkoff 7 by discussing our current code as well as our plans for our robot design.
- Shawn and Piper attended a wedding, and hence look incredibly snazzy today.
January 17th, 2011
Lab: Our robot is finally a Real Thing now! It's assembled and shiny. We had to make some small modifications/redrilling. It looks like for our scoop, we'll be attaching the motors to string, having drawbridges for both our scoop and our open door. The Plexiglas might distort some of our images, so Michael will be writing code tonight to account for this.
- Mystery Hunt was fantastic! Xavier was on the winning team, and thus will be writing next year's Hunt! Shawn and Piper have finally thawed out from running around Harvard at 3am.
January 18th, 2011
Lab: We can program our servo to do what we want! We also learned to use a lathe (at the Edgerton) , and thus made nifty pullies. We placed our center wedge in, and placed it such that balls will roll out of the opening. Our teeth for our scoop is complete.
- We have figured out our strategy utilizing Portal guns. Unfortunately, GLaDOS does not exist. Yet.
January 19th, 2011
Lab: Our robot is almost done! We currently have our teeth and servo taped to our robot instead of bolted in, our IR sensors and bump sensors need to be plugged in, attach string to our drawbridge motors, and rewire our drawbridge motors. We'll probably add a spacer so that our scoop doesn't encounter enough friction to stop our robot. This should all be doable within the next day or so. After that, it looks like we'll just have debugging ahead of us (haha, "just"). We need to recalibrate our colors and integrate our new sensors into the code.
- Piper's love allows aliens to cure cancer. Do not question X Files logic.
- Did we ever mention our most dastardly winning scheme? Teams don't seem to change their laptop passwords, so it'd be awfully easy to ssh in and...
- Our robot's particularly ferocious appearance have led us to finally decide on a name... Steampunk Cthulhu. (But the OrcBoard's name is still scully.)
January 20th, 2011
Lab: We are no longer using tape to keep our servo and teeth and various other parts together! We have lost our ghetto! We remachined a few parts and got some new ones (like our servo attachment) at Edgerton. Our servo is finally attached and done. Our drawbridge is ACTUALLY FUNCTIONAL!!!1!!eleventy!!! Our drawbridge uses pink strings, which only make Steampunk Cthulhu more fierce. It turns out gorilla glue is not sufficient for our teeth, so we're currently using tape. Yay, ghetto again!
[x] Redo teeth?
[x] Latch down collector plate so that the motors don't start drawing itself up instead of the scoop
January 24th, 2011
OH MAN LAST-MINUTE STUFF NOTHING COULD POSSIBLY GO WRONG!
Lab: Today's a day for odds and ends, in that the bulk of our robot is built and we're attaching our last (but crucial) bits. Our teeth are now bolted to our robot. Our collector plate will be bolted to the rest of our robot so that it does not raise when the scoop is supposed to raise.
- IN THE NAME OF COMPASSION AND MEDICAL SCIENCE, I CAN SAVE MANY LIIIIIIVVESSSS IF YOU GIVE ME ONE MAN. (Jekyll & Hyde is coming to MTG soon!)
[ ] Make sure to tighten all our bolts before we turn in our robot
[ ] Attach bump sensors and IR to OrcBoard
[ ] String/motor relationship should be improved for consistent winding.
[ ] Scotch tape to smooth out teeth (or a stop for our scoop?)
[ ] Maybe something for our balls to slide off of when released, to make sure they go over the wall
[ ] Do we want to play with a gyro?