Team Fifteen/Journal
From Maslab 2006
| Maslab teams |
|---|
| Team 1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · 10 · 11 · 12 · 13 · 15 |
| Team Fifteen's Journal · Paper |
Dec 27, 2005
RJ discovers the maslab wiki and putzes around with it a tad...
Thus far T15 has spoken vaguely about the project, done some research on vision crap, and set up a CVS repository for some ... theoretical pre-coding for the contest, whether or not any code will actually appear on the rep is a function of whether or not RJ is busy over break, because he's the only one with access to it, ironic eh?
Jan 9, 2006
Rob, Conor and Zack complete checkpoint one in lab. Later on they meet and discuss basic strategy and robot construction ideas. Ideas they consider are first in last out ball carrying and the general idea of "shooting" the balls out towards goals rather than pushing them through. They think these are both good ideas. Lastly, they are considering using an optical mouse over the ground instead of (or in combination with) optical wheel encoders. This would allow for (hopefully) more accurate dead reckoning. Lastly, we discussed how it seems the MASlab staff really wants teams to utilize the mapping round this year, and how we might do just that.
Jan 10, 2006
"What Hath God Wrought".... words men have only dreamed would appear on a multi-line LCD display hooked up to an I/O board with instructions coming from a super-small x86 machine. Little did the telegraph operators of the past know that on January 10th, 2006, these words would print on the screen as our pegbot, now holding on to its very existence by a few shards of duct-tape, as it approached and successfully avoided walls. Today God has Wrought us a robot that will back up to a given distance from an obstacle after detecting one ahead, then turn until it sees a clear sight. In addition, it prints data to the BotClient remotely, telling it that it has detected an obstacle, and posing the age old question to the Athena terminal operator: what HAS god wrought? After these small steps forward, Rob, Conor, and Zack got together and discussed hardware design. We have good ideas about the so-called "circular stack" to hold the colored balls in a first in last out configuration. Best of all, there will be a servo-controlled "whip" to "smash" the balls out of the holding container and into the goal. Thus, one could truly say we will be "Smashing the Stack for Fun and Profit"... well, now we need to work on some monetary prizes from the Maslab staff.
Jan 11, 2006
"Examining playing field for spherical solids with red tint!" Trumpy bellowed as the sock-clad team members jumped out of the mock arena. You can't argue with a robot, and for good reason. Trumpy has been growing big and strong as the days go by. Today we taught him what a tasty red ball looks like. Conor finessed a computer model of our chasis and ball repository while Rob and Zack crafted the makings of a searching procedure. It's too bad that Trumpy's vision processing methods are slow because the team had to scale-down the size of the captured image; Trumpy blames Rob's laziness and is apt to give him a quick kick in the shin when he's not looking. Nonetheless, Anders witnessed a miracle of life as Trumpy rotated about his wee legs and wiggled his snout before stopping dead center on a red ball. The image channel in botclient showed the silhouette of a hand moving the ball off to the side. "Silly Conor!" Trumpy teased and turned to face his prey.
Jan 13, 2006
AFter two days of furious coding, the robot now finally recognizes bar codes and reads them. ZacK and Rob worked on getting some wandering and ball chasing code together, while Conor is trying really hard to get a working robot body going. Hopefully they will catch up to their missed checkpoint on monday.
Jan 16, 2006
No lab, but intense coding on RJ's working on his robot framework.
Jan 17, 2006
RJ coded for a few hours before the lecture. Today we saw the Mapping lecture and then hit up the lab. RJ worked on his code and futzed with the robot... he has tentatively named it RowBot ... because he's obsessed with rowing. Conor and RJ were baffled as suddenly the lighting changed or God cursed us and our robot seemed to have become colorblind. To get back on track they had to recalibrate some of the values and get things in order.
Maybe the robot was just low on battery.. but it seems our motors are way too weak for use on the carpet.. .and on the tile floor they slip and can't get traction. I suspect it has something to do with our bum caster behind the bot... but who knows..
Today Conor finally got his cad model of the good chasis (as opposed to our current, ugly and cumbersome one) in a file format Rob could use and we got the base of our new chasis made in a fab. Its nice, light, and shiny. yay..
It's still difficult to get the entire team into the lab at once given that half the team is taking classes other htan maslab and some have meetings and other obligations to attend to.
Jan 18, 2006
We all gathered to discuss sensors and navigation.
RJ worked on code, while Conor mounted camera, orcboard, eden, and motors on our new blue chasis. The name of our robot has finally been unveiled to the world. It is Buffy.
Zack worked on wiring the gyro and IRPDs. He also made a couple bump sensors.
Rob had spent the prior wee hours of the morning painstakingly cutting out the new robot body.
Jan 19, 2006
We didn't make an appearance in lab today, we do just fine back at East Campus... plus our computers are faster than the athena machines in lab.. so Conor got some chasis stuff done... RJ did some code tweaking... particularly narrowing down the hues for our colorspace... We figured that running it down the hallway on 2nd West what with all the murals is a good noise-filled environment.... but just the same... Buffy performed decently well, travelling down the hallway... when a ball is sighted she drives straight up to that sucker and stops... ponders it, then proceeds forward to take hold of it for later (demonstrating CP2 and CP3 functionality) .... unfortunately directly after she tends to wander right into a wall considering ALL OUR IR SENSORS ARE CURRENTLY FREAKING DISFUNCTIONAL! (Although, thanks to zacka, they have a ::hot:: mount)
The robot falls apart and RJ dives for the off switch as a nut comes off a screw ... threatening to short some unsuspecting component on Eden/Orc... crisis averted.
Zack and RJ were heavily distracted by a slew of reuse and free food posts today, but nevertheless..
RJ and Conor head to the ITA Dinner where they supped til they had their full. (That was some good food) ITA Software seemed pretty impressive, they seem to be tackling some very difficult problems.
AND RJ codes late into hte night.. things are looking up
Jan 20, 2006
Saw the swarm demo, it was awesome. Spent most of day in lab working on multiple ball recognition and getting our IR sensors worked out. Maybe we wont destroy anymore this time *smacks hand into face*.
Currently working hardest on getting our PID controller to work...
Jan 21, 2006
Bad Ideas at East Campus..... no work got done
Jan 22, 2006
Still bad ideas.... but some work on the gyro got done...
Jan 23, 2006
Hitting hte lab early today... RJ starts his 2nd IAP class today so now most of our members have conflicts during the early afternoon.. but hopefully well be able to all get in there after that... RJ spends the morning coding and in lab after he gets out of crew.
Jan 24, 2006
Conor Finally finishes re-wiring the new IR sensors, triple checked for shorts and to make sure the pin-outs were right, and got them ready to go back on the bot. He then built a little gear box for the high speed motor so we can start sucking and shooting balls into our bot. RJ codes HARDC0R3 all day, except for when he is at 6.balahblah something or other. I (conor) think he is perfecting the robit for checkpoint 2/3/mock competition, and it should be ready for that soon. As Soon as conor figures out how this durn motor gizmo is going to attatch we are going to move to version 3.0 of the body, which Conor thinks should be called Virgin IR Sacrifice (or don't put 5 fricking layers of heat shrink over a bad soddering/wire connection job). Could be shortened to vIRs, though that doesn't make much sense at all. So it goes. Also, Rob Built a metal camera mount so that the camera doesn't constantly shift within a duct tape mess (though Conor misses the presence of duct-tape greatly, and will probobly start duct taping random usless parts onto the robot just for the sake of visual effect). Zack probobly did something today, but this reporter doesn't know what that something is, so he will go unmentioned for now.
Jan 25, 2006
Jan 26, 2006
Jan 27, 2006
Jan 28, 2006
RJ coerces Buffy into wandering around 2nd West. It's good; real good.
Jan 29, 2006
RJ magnanimously tweaks and optimizes code while the other three take a short day-trip to Mars (Conor goes AWOL. Zack & Rob discover gold on Mt. Liberty, NH)
Jan 30, 2006
The final push toward glory: all team members busy themselves on a strict schedule. RJ and Conor use the early afternoon to finish the roller. Rob sits in front of the design and ponders servos. Zack is thinking about bump sensors Also, the Robot has an official name change today, no longer Buffy (seeing as Conor never got that joke anyway) it is now called the Virgin IR Sacrifice, to commemmorate our team being stupid, and dumb... and stupid again.
Jan 31, 2006
Feb 1, 2006
RJ engineers the ultimate balance of finite states and tunes goal-finding and scoring behaviors. Eden's hard drive gives him some grief.
Zack fabricates tactile sensors for the front of the robot. Suddenly Buffy doesn't act so stupid when it hits a wall.
Rob spends the entire day constructing front roller assemblies. Three motors later, the next day, he fires a ball from three feet away into a goal. Five minute epoxy is better than eight hour epoxy.
Feb 2, 2006
RJ continues onto the next day with morning crew. Buffy misses him. Rob executes the bot code with the new roller. Buffy picks up both balls and scores them.
