02-03-2011 22:12

Akash Rastogi

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Love the song, love the robot as usual.

Favorite robot so far. That manipulator is just epic and efficient as hell.

02-03-2011 22:15

Andrew Zeller

Re: pic: Killer Bees 2011 Robot - Buzz XVI

WOW. That is an amazing and extremely effective design. Team 33 will go very far this year.

02-03-2011 22:17

Dancin103

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Another year, another great Killer Bees robot. It looks awesome and good luck this season!

02-03-2011 22:18

ATannahill

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Good looking robot, I'm looking forward to seeing it this weekend.

02-03-2011 22:19

Vikesrock

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Looks like another fantastic robot from the Killer Bees.

The two tube automode looks pretty sick. Is there any vision or other sensing being used or is it all dead reckoning with encoders?

02-03-2011 22:19

Nuttyman54

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Was that a 2-ubertube autonomous mode? Kudos to you guys, maybe the programmers won't look at me like I'm crazy when I tell them to go for it...

02-03-2011 22:22

Eric Yahrmatter

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Dead reckoning with encoders and a gyro.

02-03-2011 22:22

arob9119

Re: pic: Killer Bees 2011 Robot - Buzz XVI

wow, great robot guys! This is definitely one of my favorites so far.

02-03-2011 22:31

Akash Rastogi

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Stealin from the Foley playbook.

02-03-2011 22:33

nuggetsyl

Re: pic: Killer Bees 2011 Robot - Buzz XVI

dammmmmmmmmmmmmmmmmmmmmmmmmmm

02-03-2011 22:39

pwnageNick

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Looks stellar. 2949 will be watching you on the webcasts this weekend. Can't wait to see you guys in St. Louis.

02-03-2011 22:41

nikeairmancurry

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Your not =D

02-03-2011 22:47

Jacob Paikoff

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Wow

Great claw, the over the back feeding ability is so cool.

Another great Killer Bee robot

02-03-2011 22:51

jblay

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Everything about what I just saw was awesome. Its so very awesome.

One recommendation. I would put red tape or something around the border of the mini bot centerer plate because that can't be easy to see from the other end of the field in the middle of a stressful match.

This video absolutely made my day. Hope you guys have another great year.

02-03-2011 22:54

apalrd

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Since the filming of the video, we colored the rollers on the end yellow and the supporting poles black. Now they're antennas!

03-03-2011 07:12

Josh Fritsch

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Looking great guys, see you today at Kettering!

03-03-2011 07:28

Alpha Beta

Re: pic: Killer Bees 2011 Robot - Buzz XVI

We had a 2 uber-tube auto planned, but there was a spin move in the sequence which cost us some time. You may have inspired us to put the concept back on the table. 15 seconds might be just enough time afterall.

I love that you can do the auto with just encoders and a gyro. Veteran programming teams ought to be able to help those without autonomous get something together during Thursday if they have the manpower to spare. What do you use as sites to line the robot up initially?

Beautiful robot. You've inspired us once again.

03-03-2011 08:32

IKE

Re: pic: Killer Bees 2011 Robot - Buzz XVI

For the first tube:
The 2 minibot towers are not only a rule, but a useful alignment aid. Find something on your robot to align with the towers, and you will have a constant depth reference. Then make sure it is pointed straight up the field at the goal you are intending.

The second tube requires a bit more creativity (but not much). It does require a lot more patience and tuning. Make sure you get 1 and get it repeatably before going for 2. Never compromise 1 for 2. Doing 2 requires a partner letting you, which can be more difficult than you might think. We had a 5 ball auto last year that we were only able to use a couple times.

03-03-2011 08:39

Rangel(kf7fdb)

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Just have to say, beautifully robot. Looks like we are not the only ones who can slam into the feeder station wall to get a tube without braking our arm.

03-03-2011 09:15

GaryVoshol

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Make sure that "something" is the forward-most BUMPER per the Rule <G04>.

03-03-2011 09:22

IKE

Re: pic: Killer Bees 2011 Robot - Buzz XVI

<G04> Each ROBOT must be positioned on the FIELD so that the BUMPER closest to it‟s ALLIANCE‟S PLAYER STATION breaks the plane formed by the POSTS of the opposing ALLIANCE‟S TOWERS. Violation: The MATCH will not start until all ROBOTS are in legal positions. A YELLOW CARD may be assigned for repeated violations.

Interesting. We took this to mean that the bumpers had to break the plane thus you would have approximately 1 robot length of wiggle room to adjust. Just to understand, your take is that it is the front (and only front), and thus you have about 3" of wiggle room. Correct?

03-03-2011 09:29

Josh Fritsch

Re: pic: Killer Bees 2011 Robot - Buzz XVI

I also was under the impression that as long as the front bumper is closer to the human player station than the plane of the tower posts (thus breaking the plane) you would be good, giving you approx 1 robot length of wiggle room....but I guess we will find out soon enough.

03-03-2011 09:41

GaryVoshol

Re: pic: Killer Bees 2011 Robot - Buzz XVI

(emphasis mine)
My interpretation is that this means that portion of the BUMPER that is closest to the STATION must align with the POLES. There was another thread on CD about this. I don't recall if it was in Q&A or not; I'll look before tomorrow.

03-03-2011 10:45

Chris is me

Re: pic: Killer Bees 2011 Robot - Buzz XVI

With robots like this, I can't believe you guys have a bad luck streak going since 2005.

This is the year you'll get gold, I guarantee it.

03-03-2011 11:01

Katie_UPS

Re: pic: Killer Bees 2011 Robot - Buzz XVI

I'm going to miss seeing all the witty signs from hangs

03-03-2011 11:12

Tom Bottiglieri

Re: pic: Killer Bees 2011 Robot - Buzz XVI

I was wondering how long it would take for someone to post a 2 tube auto. Good stuff! I think we will see a few teams (my guess: 5) who can do it this year.

03-03-2011 14:33

BJC

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Something that isn't really highlighted in the video: Our drivetrain.

Our drivetrain this year is the first of its kind. Think 254 + 148.

03-03-2011 14:41

Akash Rastogi

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Yes I did hear something about an optional 8wd.

03-03-2011 17:25

1086VEX

Re: pic: Killer Bees 2011 Robot - Buzz XVI

2 tube autonomous equals pure win!!
im lovin the bot! very clean, unique, and effective!! cant wait to see it in competition!!

04-03-2011 10:39

JesseK

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Fun video, awesome song! I like how your sponsors are right up front. They will like it too

04-03-2011 10:56

ktress357

Re: pic: Killer Bees 2011 Robot - Buzz XVI

I like it!

04-03-2011 22:01

GaryVoshol

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Two double-autonomous scores at Kettering today.

04-03-2011 22:44

The Lucas

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Did the second one in match 32 count? From the video it looks like they were given a penalty for hanging the second one in teleop. I didnt hear an explanation since I had that webcast muted (audio was choppy). Either way they got it clean the next time.

Big congrats to 33 (and 148) for their uber auton skills! I didn't think we would see a single double uber autons in Week1, let alone 4 by 2 different teams.

[EDIT] I somehow missed match 36. I forgot that Districts run tight schedules. Gary was right, 2 double ubers today [/EDIT]

04-03-2011 22:53

BJC

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Unfortunatly, that double did not count and was instead a penalty. What happened was that when our claw tried to open it was at the perfect height so that when it tried to drop the tube it couldn't move down as it somehow caught on the top of the driver station. Either way, the auton was fixed after the match and the next two yielded better results. Looking forward to tomarrow!

04-03-2011 23:17

kgzak

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Saw this thing in person today and it was amazing! I'm scared to see you at West Michigan (when we are actually competing)

05-03-2011 21:51

AllenGregoryIV

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Wow, we built a holonomic drive train so we wouldn't have to slow down to spin around form the feeder. You guys just went over the back of the lift, that is such a cool design feature and something I never thought of. Does any part of the collector stick past the bumpers when driving with a tube and are you worried about it getting hit?

The double ubertube is really impressive thought I wish more teams were trying to do more than just dead reckoning. They give you so much to sense off of, why not use it?

05-03-2011 22:24

apalrd

Re: pic: Killer Bees 2011 Robot - Buzz XVI

FYI:
My double-ubertube code uses only what most people would consider "dead-reckoning", using only encoders and a gyro for positioning.

We do not have a camera or any line following sensors at all.
During testing, we have never missed the first tube if the base drive functions are correctly calibrated.

05-03-2011 22:46

AlexD744

Re: pic: Killer Bees 2011 Robot - Buzz XVI

I personally have found "dead reckoning" to be very reliable and must quicker to program. Last year, for example, we were able to score both mid-zones balls every match with "dead reckoning." This year we have already tested scoring on both the top and middle rows with "dead reckoning." It works consistently as long as the robot is lined up the same way every time (It's nice having a programmer on the drive team). Considering the time constraints on testing autonomous, its much faster to get a working auto using this method, at least in my opinion.

Also, I was playing around with the idea of a double tube autonomous before ship, but decided against it. Now that I've seen it work, I think I'm bringing it back to the table. It'll take some doing, but I think I can get it working, if not for Florida, at least for South Carolina. Thank you 33.

05-03-2011 22:51

BJC

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Neither our over-the-back move or our carry position stick out of the bumper zone.

It's probably worth noting that we never even went into our lane in any match (never needed the over-the-back feature). It was simply unnecessary with the amount of time it takes to traverse the field and the availability of tubes on the ground.

05-03-2011 23:11

Lil' Lavery

Re: pic: Killer Bees 2011 Robot - Buzz XVI

What, if any, sensors do you use on the elevator? An encoder?

05-03-2011 23:43

Brian Ha

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Lol there bot was just amazing ovver all, they were great. Loved playing against you guys and with you when the chance happened. Espically in the semis. We got jacked for place though, if u guys didnt hear what happened, for all those teams out there make sure you pass inspection otherwise, stay in the pits dont come out! It will make it worse for your alliance when you all get dq. Please and thank you.

06-03-2011 13:33

AllenGregoryIV

Re: pic: Killer Bees 2011 Robot - Buzz XVI

I wasn't trying to down play dead reckoning, I completely understand why teams use it and I know that it is extremely effective. I was just saying that there are other ways to do it and I like it when teams try to figure it out. This year we were awarded the Innovation in Control award at Alamo for using two Sonar Range finders to position our autonomous mode against the lane divider and driver station wall (the initial plan called for 4 but we never got that fully working). The sonars make autonomous much easier because all we have to do is make sure the robot is strait, it's position behind the line doesn't matter at all. The plan with 4 wouldn't have needed the robot to be strait because it would have corrected it's orientation against the front wall. We also haven't been able to pull off the double ubertube but that is mostly do to lack of testing time with the robot so maybe dead reckoning is a better solution.

06-03-2011 13:56

apalrd

Re: pic: Killer Bees 2011 Robot - Buzz XVI

10-turn pot for the elevator, 270degree pot for the wrist.

During auto, all of the elevator commands are fed to the normal state-machine that is always running. The auto code only handles the drivetrain directly, using two encoders and a gyro for drivetrain positioning.

So, by "for positioning", I meant "for drivetrain positioning".
That's what I get for posting at 11:00 after two long days of competition.

(I actually write the auto routines in beescript, and use the beescript interpreter to call the auto command blocks, which set the state machine or drive, but that's mostly to avoid LV downloads when making auto routine changes)

06-03-2011 14:22

tr6scott

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Awesome Robot, and Awesome Alliance, thanks for picking us.

So can you send me a link or info on beescript, or how you do state machine stuff, and avoid the LV download.

I am a first year mentor for programming on team, and had my hands full with my first exposure to LV. Having come up through the FLL program, and placing 4th and 2nd in the state a few years back, we have that line following down.

Thanks for the help, and kudos to you and all of the BEES, 33 rules.

06-03-2011 14:25

Lil' Lavery

Re: pic: Killer Bees 2011 Robot - Buzz XVI

I kinda figured you meant drive positioning. I was just curious about your elevator sensors, as we had some issues (which in large part, where a result of us not truly planning ahead for sensor mounting) with our elevator sensors and had to get creative in terms of how to measure it. Just wanted to see what you guys had done.

06-03-2011 17:42

apalrd

Re: pic: Killer Bees 2011 Robot - Buzz XVI

1. You guys are awesome too.

2. Beescript. I wrote it myself (and named it after Bees, of course). Basically, I read a text file on the cRio (which I upload via FTP), and separate it into an array of lines, and then do this:
If the line starts with a #, skip it (it's a comment)
If the line has <3 characters, skip it (it's probably a space or blank).
Else, tokenize by spaces into an array of words, lookup the first word in an array to find the index, then look through another array of commands to find the command ref. Call it by ref, with the rest of the array as the argument (you have to use strict-type VI refs to do this).
Each command has the argument array input, and then looks through the arguments to find the ones it wants(usually by assuming the first argument is X, the second is Y, etc.), and uses those to run what it runs. Upon entering auto, it populates the array, with each command having the command name and the ref in an array (look at the WPI refnum registries for an example of how to do this).

3. State-machines. I have an array of cal points, each of which has an elevator height and arm position. I also have an enumerated type which defines the state. I double-buffer the state to handle going over backwards and forwards, since those use collision avoidance. I take the resulting state and lookup the position in the array (since enumerated variables can be used as unsigned int's) and also have a few special states which I further process (especially the score positions, which have score sequencing). I have a single get-set VI which handles state storage, double-buffering, and initialization. If it is disabled, it will uninitialized, and a command must be sent to re-initialize. The motors will be set to 0 if it is not initialized, this is a safety feature.

Simple: I use an enum and an array of positions, then lookup the position in the array. I also have a large case structure for special cases.

3a. I also have a few VI's that run after the state-machine, one which actually does P control to determine motors speeds, another which prevents the window motors from stalling (we have a few states which hit mechanical limits, so we drive the setpoint past the mechanical stop then kill it after it stalls to prevent PTC-tripping), and a third handles manual override.

3b. Since we have two robots (competition and practice), I translate all of the sensor inputs into inches and degrees, so I can use a single set of calibration data on both robots.

07-03-2011 11:21

tr6scott

Re: pic: Killer Bees 2011 Robot - Buzz XVI

Ahhhh... you are a most clever code monkey...

that is much more than I am willing to take on at the moment, but we do have a few key variable data that needs to be tweaked and tested. I will look at what you have provided, and try to coach up the team.

I have done this in the PLC world with servo controllers, where the servo is programmed to do one move with variable data, and I just adjust the data as required.

Can you point me to where the data file should be placed on the cRio directory structure, and what type of file to create... standard text I am guessing is ok?

I will google around later today.

07-03-2011 12:54

apalrd

Re: pic: Killer Bees 2011 Robot - Buzz XVI

If you FTP to the cRio, you can pick your directory. I put mine in /ni-rt/startup but the path in LabVIEW is c:ni-rtstartup for some reason.