Meet Team Antipodes: All-Girls Robotics Team Shares Their Innovative Robot Design


Uploaded by GoogleTechTalks on 21.06.2010

Transcript:
>>
CHIPPINDALE: Welcome, everyone. I'm Kjersti. >> REPLICON: I'm Violet.
>> FILAR: And I'm Emma. >> CHIPPINDALE: And we're team Antipodes,
an all girls team from Pacifica. We recently went to an international competition in Istanbul,
Turkey and competed there. And today we're going to be sharing with you some of our experiences
there as well as other tournaments, and also, a little bit of our robot and other aspects
of FIRST LEGO League competitions. >> FILAR: So the first tournament we went
to was in Santa Clara. There weren't very many people but we still got first. So, we,
from there, went on to a qualifying tournament in Sacramento which was a bit larger and there,
we took third place which qualified us to move on. And then we made it to the Northern
California championships where we got third place again and we were able to move on to
the open European championships. So they were scheduled in mid-April but the volcano erupted
and covered Europe with a giant ash cloud, so it was postponed.
>> CHIPPINDALE: Wait, wait, wait, wait, it's called Eyjafjallajokull.
>> FILAR: So, just last week on the second and third, we participated in the tournament
in Istanbul with over 50 teams from all over the world. We did not place but we still did
very good. >> REPLICON: Okay. So you may be wondering
how this got organized and such. So FLL, FIRST LEGO League, is the competition that we participated
in and it is--it is one of four that is organized by FIRST. FIRST is an organization sponsored
by Dean Kamen, the inventor of the Segway, and it stands for "For Inspiration and Recognition
of Science and Technology." There are four programs organized by FIRST, the first one
is Junior FLL. It's for ages 6 to 9 year olds, it also uses MINDSTORMS robot. The one we
were in FIRST LEGO League is for 9 to 14-year-olds. The next one is FIRST Tech Challenge, this
uses a bit more complicated VEX robots which are metal pieces that you put together and
that's for 14 to 18-year-olds. And then there's FIRST Robotics Challenge which you basically
make your own robot and that is also for 14 to 18-year-olds.
>> CHIPPINDALE: So, on the course, the robot remains completely autonomous and so if you
touch it while the robot is outside of base, you get a 10-point deduction from the 400
total points you can gain from doing all your runs. There are 11 loops set up around the
course like this and if you bring--each one that you bring back to base is worth 10 points,
so you can get a total of 110 points from getting the loops. But there's also a bonus
involved. If you get three grey loops, you get one free red loop and if you get all three
red loops, you get one free loop of any color. So, that means you don't have to physically
bring all the loops back to base, you get some of them for free. If you release the
car from the ramp right here, you get 20 points. If--there are access markers in the corner
and if you knock each of them down, they're each worth 20 points, so you can get a total
of 100 points. If you end on either the top of the bridge or the target area, you get
25 points. There's a crash test dummy that if you just kept in or on your robot the entire
time, you get 15 points. If there--you have four passengers and if they end in the target,
you get 10 extra points. There are also some obstacles. In the corner, there are four red
walls and if at the end of your run they're all standing, you get 10 points for each wall,
so a total of 40 points. And then there's also access or warning beacons that are set
up all over and if by the end of your run they're all standing, you get 10 points per
beacon, so you can get a total of 80 points. >> FILAR: So each robot has one CPU, which
is the brains and helps control everything that it does. You're allowed to use one ultrasonic
sensor, up to three motors which have built-in rotation sensors, up to two touch sensors
and up to two light sensors. Along with this, you can use any other product that is built
by LEGO. >> REPLICON: Okay. So now that you know all
the tasks that this robot has to accomplish and all the parts that are allowed in the
robot, I want you to think of a robot that you would design, that would reliably get
400 points every run in two minutes and 30 seconds, which is the time limit for FLL.
So close your eyes. I want everyone to close their eyes and I want you to imagine for 30
seconds what you would make. Okay, so not exactly 30 seconds. So I want--so as--I know,
as probably all of your ideas are very different from each other, similarly, all the FLL teams
have very different robots. So here's one of them and as you can see, they all have
different attachments, different strategies, and they're completely different. So here's
our robot. >> FILAR: No.
>> CHIPPINDALE: No. >> REPLICON: No, there's our robot. So, one
thing unique about this robot is that it's very simple and its attachments. We want to
spend as little time and space as possible and more time out on the course completing
the tasks. >> FILAR: So we have a touch sensor with a
plate so that we don't have to actually touch the sensor. We have two light sensors, rear
guide wheels so we can run along the wall. The CPU frame on the bottom to hold it together
and the block extra light, a front touch sensor, two drive motors, two drive wheels, two front
wheels, another set of guide wheels, a arm and axle--I mean, a motor and axle for our
arm, sides, a front frame, a support for the motor, a roof, and a sensor activation plate,
and an arm. And this is it together and blown up.
>> CHIPPINDALE: So after you've done designing the robot, it's time to program it. There
are--we use a programming language called Mindstorms and XTG which is an icon-based
programming language that is much easier for younger users to learn. And there are simple
blocks such as the move or motor block shown here that will just make the robot go straight,
turn or something like that, but there's also more complicated arrangements of blocks such
as the line follow program shown here in which they use a lot more complicated blocks. And
so, once you get into more complex programming, you can use loops which the line follow is
contained in, if then switches, and you can also use your programming to orient yourself
around the course. So, for instance, you can back into a wall and if--when the touch sensor
is touched or pressed, it will stop the robot and so that way you know exactly where you
are on the course in relation to the rest--to everything else. And our robot does that a
lot of times and so now we can show you just what it does. So we're going to run our robot
for you now. Usually, before we do runs, we do a light calibration because during the
line follow program, we do a line follow or we do--we use the light sensors. And if the
lighting conditions will change and so if we just had one slight value for all of them,
it might mess up, and so we have a shot on the bottom as well as doing a light calibration
in which it reads the value of white and the value of black and our line follow program
finds the average of those two and stays on the grey. So, it stays right where the black
meets the white instead of only on the white or the black.
[PAUSE] >> REPLICON: So since now that we got all
three grey loops, we can get the red loop, and since we got all three red loops now,
we can get the blue loop right here. Oh. So sometimes, our robot can be a bit unpredictable
like that instance right there. Sometimes, the sensors are a little bit off, so it'll
run into something and it'll have a collision like that.
>> CHIPPINDALE: So now in a real competition--so now in a real competition, the referees would
come over and they'd review everything with us and then we'd get told how many points
we got. So, first, we start with the loops, we got 10 out of 11 loops, so that's 100 points.
We got the car, so that's 20. We got off our access markers, so that's another 100. So
we're up to what? 220? And then, we actually did not land in the target and--but we did
have busters, so that's 15, so that's 235. We did not knock over any access markers,
so that's 315. We also didn't knock over any walls, so that's 355 and--but we didn't land
in the target and we didn't land with the people in.
>> REPLICON: And [INDISTINCT]. >> CHIPPINDALE: I said that. And so that is
our run. >> REPLICON: So this whole experience of doing
FLL, the FLL season, has been an adventure and quite a learning experience actually.
We now know more about how the engineering process works, how to make our ideas become
reality, and how to work as a team. It has also helped us decide what careers we want
to pursue in the future. If you know anyone, adults or kids that are interested in FLL,
please let us know because it's such a great experience, and there's a lot of kids out
there who would love to be a team but they just don't have a coach to get them started.
So now, we invite you to come up closer, we can run again. If you have any questions,
we can answer them and you can take a closer look at our robot. Yes?
>> Can you elaborate on your career plan? >> REPLICON: I am--personally, want to be
a robotics engineer. I would love to work for NASA one day, build satellites and rovers
and such. What do you want to be? >> FILAR: I want to be a vet.
>> CHIPPINDALE: And in doing this, you have to present a lot and teach what you've learned
to a lot of people as what we're doing now. And through this experience, I've learned
that I definitely want to become a teacher. >> REPLICON: Any other questions? Yes?
>> Challenges of building the robot [INDISTINCT]? >> REPLICON: One of the big challenges of
building it is that we chose to go under the bridge here instead of--let's see if I can
go over here. >> Other teams went under...
>> REPLICON: You can go under here and cross over these. And it's--we decided to go under
here, so our robot had to be a certain height and also over here in the corner. Kjersti
wants to point. There's a very small space we can turn in, so our robot can only be that
wide. So our robot is as big as it could be to do the task as we have, and also programming
line follow was difficult. >> [INDISTINCT]
>> CHIPPINDALE: So the line follow we did was called a PID line follow program which
is proportional, integral and differential. And so basically what it did was at the beginning,
as I said, it--we have two light sensors and one of them would be--the light sensors detect
how much light is reflected back off of the mat when it shines the light on it and so
there, you assign numbers. And so, for instance, white would be 50 and black would be 100,
except for it'd be probably be black is 50 and white is 100. And so, what it would do
is one light sensor would read 100 and one would read 50 and so instead of staying only
on the white and only on the black, it would stay--one light sensor would be on the black
and one would be on the white. And so really, it would be staying right in the middle, so
it'd find the average and stay right in the middle the whole time. But as you saw, it
it's not--it's easier to say than to write the program for and it's a very long program
and it's--most of it is math and calculating. We had a lot of variables that we had to make
which are set numbers that are repeated throughout the program. And it's all calculating, calculating,
calculating and at the very end, there are about three move blocks, and so it just repeats
that over and over again to do the line follow. That took me about two to three weeks to write.
>> REPLICON: Any other questions? Yeah? >> [INDISTINCT] when you talked about how
the robots from different teams used radically different routing each other from yours, when
you guys first started up to design your robot, how much agreement or disagreement or a variation
was there between the three of you in terms of building one robot?
>> REPLICON: Well, there is--I was basically the only builder but they definitely had input.
I think the main differences we had was in the strategy which definitely influenced how
the building of the robot went. And also, in building the arm, one of the things about
the arm is it stays parallel to the ground the whole time due to the--there's a stable
gear and the other gears rotate around it. At some point, we wanted--a few of us wanted
to just have an--just an arm that stuck out and just went like that and others wanted
the parallel, so there's definitely disagreement in that. Any--yes?
>> You mentioned what's the variation [INDISTINCT] well, you designed your robot and then you
go to [INDISTINCT] but I'm curious as to what the design process and its interaction with
the programming process, how does that--usually these things tend to [INDISTINCT] everybody
around. >> FILAR: Well, it's--we kind of--we had to
get the basic build of the robot and then we started programming, but then at the same
time, we would be building more and changing things, so then we would have to change the
programming. And it's a little bit difficult especially for the programmers. Well, I mean,
it's difficult for the builders too, but the programmers will write a program that they
like and it's all working and then the robot will be different and they have to rewrite
it some other way and so it's a little bit difficult to just keep going through that,
but in the end it works. >> You obviously [INDISTINCT].
>> CHIPPINDALE: Both of us. We had a lot of time constraints as a team. She was doing
theater, I was in marching band, and Emma was on track. And so, we all had after school
activities and it was really hard to meet as a team and so, most of the time what we
would do is we would meet as a team on Monday and we would decide as many strategy decisions
as we could make. And then we would--the program, we would program and then the designers would
meet--the designer would meet and she would design. And--because what would happen if
we were all together is that she'd be trying to build the robot and we would be trying
to write the program and you can't do both at the same time. We'd be fighting for custody
of the robot. So we ended up working--we were working together on the important strategy
decisions but then we would separate and work independently for the building and developing.
Yeah? >> Can you talk about the layout course? When
is that decided? Is it before the competition [INDISTINCT]? When do you know it's going
to find that kind of stuff? >> REPLICON: So you--there's a main theme
for it. It changes every year. That is released like May, so we found out--well, yeah, the
theme is released in May. And then the course and the mission and the mat and everything
is released in September and they'll email out everything to you. And usually, the first
real robotics meeting you have is putting together all these pieces on the mat; they
just send you the pieces and the instructions and you have to put everything together. And
then once you have it, also that we really start strategizing, so that's all in September.
>> CHIPPINDALE: And then the first competition is in November, so you don't have a whole
lot of time. >> REPLICON: Yeah.
>> How do you [INDISTINCT]? >> REPLICON: Yes, I think. One of the things
we didn't realize you can do, for this blue loop here; we didn't want to get it because
it has beacons all around it. But what some team--the other teams decided is that, "Oh,
you can bring the beacons back to base and stand them up in your base and you won't lose
the points." We didn't think of that until we saw someone else do it. We're like, "Oh,
my goodness, that would've been great." And we've had so much trouble with that orange
loop back there that it would've been a lot simpler. I think that's something I personally
would change.
Yeah? >> The weirdest thing that happened, how did
you figure out what [INDISTINCT]? >> REPLICON: Yeah. We had our last competition,
the first day, our robot can't run straight. We were just like, "What is happening?" And
actually, two things were sort of going on. One, when the motors was completely breaking
down and another is that we had these plastic wheels on the bottom, there are front wheels
I guess there, and they were rubbing against this other plastic piece and so, they're making
these grooves and we didn't realize this 'til later. And so there's these giant plastic
grooves that we would--did not expect at all and that was completely affecting the straightness
of the robot. That was something unexpected. >> FILAR: Also, one time, I put the slide
on backwards and we realized that it actually works better, so we just changed it. Any other
question? Yes? >> How much [INDISTINCT] other grown-ups or
kids or, you know, [INDISTINCT]. >> FILAR: Do you want to talk?
>> CHIPPINDALE: We have a coach, Ken Filar. You're allowed--you can have many coaches
and you can have mentors which are generally kids older than yourself who will help you
with strategizing and everything. And the purpose of coaches is not necessarily to like--they
help you along and if you get stuck, they can give you suggestions but they try to not
interfere with, like, your--with you building it. It's supposed to be the kids, not the
mentors. And also, we--along with this in the competition, there's also a project part
where you have to build--pick something--this year, it was about transportation, so you
picked something in your community that you wanted to change about the transportation
and in that, you could get a lot of, like, outside help to--for research and whatnot.
>> REPLICON: They actually encourage you to. Yeah?
>> [INDISTINCT] how many years have you been involved [INDISTINCT]?
>> REPLICON: Well, this is actually my first year of doing this competition but I've always
loved robotics, like, just robots. I've always wanted to build something like that and here...
>> FILAR: So this is me and Kjersti's second year. And the year before, we were on a big
team that was just kind of randomly put together. We got dead last at the first competition.
I think our highest score was 40 points and our lowest score was like 20, which was pathetic.
And so, we definitely--me and Kjersti learned a lot. The first thing we did was we decided
we wanted a much smaller team with people who actually were going to work and care,
so it became us. >> CHIPPINDALE: One thing we learned from
the previous year is that the previous year, we had spend all our time in base and putting
on different attachments and changing little things that really took a lot of time out
of the mission and is part of what caused us to only get 40 points. And so that definitely
helped us this year. It was one of the first things we decided was that we were going to
spend as little time in base as we could. And we practiced over and over changing out
the--doing our base time to make sure we got just a little base time as we could because
we didn't want the same thing that happened last year to happen this year. Is there a
question? >> How much did the robot change from the
level competition changed on the course since you had testing in November [INDISTINCT]?
>> REPLICON: It didn't change that much. Actually, we have a shroud on the bottom that we built
because at one of our competitions, our line follow just didn't work and it's because there's
this big sky lights at the top and the light level would change during our run after we
did our calibration. So that's one of the major things that changed. I think basically
everything else stayed the same. A lot of programming differences we...
>> CHIPPINDALE: Getting the orange loop in the corner, we went through many, many trial
and errors in trying to get that. There was--but we ended up going back to the same thing that
we had at the first competition. So, if there wasn't a whole lot of difference but there
was a lot of progression and learning and trying to get that to work. And also, from
the first competition, at the first competition, we didn't have, in running it; we didn't have
a high score of 400. We had a high score of about 370 because we didn't get the grey loop
on there, which means that we didn't get three other loops and--or two other loops. And so
throughout that, we've gradually gotten--we've gotten more and more points also as the thing
went on. > You didn't try to [INDISTINCT]?
>> CHIPPINDALE: Yeah, yeah, or we didn't try to get that at all. We just left it out and
hoped to get 370. Yes? >> [INDISTINCT].
>> CHIPPINDALE: At the first competition, it was in Santa Clara so we just drove there
and there were about, like, 16 teams maybe which is a lot fewer. At the second, there
were maybe 37 about and at the NorCal tournament, there were about 48. And then at Istanbul,
there're about 52, I think, from all over the world. And to get to Istanbul, we had
to do a lot of fund raising from local businesses and more corporate businesses to get the money
to buy plane tickets and hotel and everything. >> REPLICON: But due to the volcano, a lot
more expenses came up, so we're still trying to fund raise. We're still short.
>> CHIPPINDALE: We built a website to help to get funding, so we have a website and you
can go to and just like $10 helps, always a good fundraising opportunity.
>> What's the highest score you got? Four hundred?
>> CHIPPINDALE: We've gotten 400 multiple times. Yeah. We didn't get any at the Istanbul
tournament, but at the NorCal tournament, we got one 400 and we've gotten 400s in practice
many times. Our robot is unpredictable, you never know when it's going to be a 400 and
you never know when it's going to be, like, a 300. Any other questions? Yeah?
>> I guess [INDISTINCT]. >> REPLICON: Me and Emma--well, actually me
and Emma, in the fall, we're going down to Tasmania because we have a sister team actually
down in Tasmania and they invited us down. So we're going to do FLL down there since
it extends to 16 in all other countries except the U.S. And then after that, I think we're
going to do FTC. We decided FRC was--you need, like, a 20 person team and we're not going
to be able to recruit that many people. Yeah. So, any other questions? Okay. You can come
take a closer look at our robot, see how it works. We can probably run it again if you
would like, so, yeah. Come on. >> CHIPPINDALE: So off to the side is our
research project, which you can take a look at which is mat gloves. We have working mat
gloves. Somewhat working with mat gloves. >> REPLICON: Thank you.