PDA

View Full Version : Image Discuss: Team 190 "INS"


CD47-Bot
02-19-2003, 12:09 AM
[cdm-description=photo]14923[/cdm-description]

Caleb Fulton
02-19-2003, 12:10 AM
Looks cool!

What exactly does it do?

Joe Matt
02-19-2003, 12:11 AM
You don't line track, you don't dead recog. You don't jump over the wall and you don't use electronics to determine where you are relative to the ramp. So how the heck does this work? :confused:

Stephen Kowski
02-19-2003, 12:13 AM
were you using that at the UTC scrim?

WernerNYK
02-19-2003, 12:16 AM
No, the system was not available during the scrimmmage.

Stephen Kowski
02-19-2003, 12:17 AM
did you develop it in between, or did you have it made, but never put it on?

WernerNYK
02-19-2003, 12:29 AM
Well, WPI has been an active FIRST participant for the past 12 years...


Nah, just kidding; it's been our plan from the start of our brainstorming. :yikes:

Gope
02-19-2003, 12:32 AM
There are two ways you could do this, that i see.

One, ultra sonic sensors

Two, built a virtual court and used motor rotations to determine change and rate of change of position.





Somehow I doubt you did number 2. So what did u do?

JVN
02-19-2003, 04:00 AM
It's part of their CVT device...
It clocks the velocities of those via IR wheel and uses that input to determine robot postition.

Jeez guys, they showed it off enough last year... Doesn't anyone remember it?

It's nice to know you finally got it working.

WernerNYK
02-19-2003, 04:15 AM
Well it's nice to know that you know how our CVTs work (afterall, we are in this thing not only to design great robots but also to teach and inspire :) ).

But sadly enough, that box there has nothing to do with CVTs, mainly because there are currently no CVTs on the robot. :eek: But concerning those CVTs from last year, we did get "that thing finally working" for next-gen CVTs. :D

nkrumm
02-19-2003, 05:16 AM
I'm pretty sure they are using inertial sensors (i.e. an accelerometer) and integrating the value they receive to calculate their position. This would allow you to get quite accurate results, better than encoders, since the accelerometers can pick up wheel slippage and shoving.

I'm quite interested in knowing how you account for overall "drift" in the system... (?)

:: Nik

nkrumm
02-19-2003, 05:16 AM
I'm pretty sure they are using inertial sensors (i.e. an accelerometer) and integrating the value they receive to calculate their position. This would allow you to get quite accurate results, better than encoders, since the accelerometers can pick up wheel slippage and shoving.

I'm quite interested in knowing how you account for overall "drift" in the system... (?)

:: Nik

sevisehda
02-19-2003, 10:56 AM
Military has used inertial guidance systems in the past. Subs used them to navigate under the ice caps and in deep water, before the days of GPS. ICBMs used them to. The only problem is they slowly lose accuracy. They would also have to put the robot in the exact same orientation each time so it would know where to start. I doubt it would be very acurate after a hard hit because most accelerometers only work to a certain point.

Brian C
02-20-2003, 11:00 PM
A good point!

The system may end up having some drift but I'll bet that those 190 folks have their act together when it comes to the overall package.

They've been building 'bots in Worcester for a whole lot longer than many teams have been around and experience ALWAYS helps......

Team #311 won't be competing against them at any of the Regionals...but we'll see you at Nats (again) and maybe even at BattleCry4.......

Hey Chris.....tell Matt C. I said Hi!

Deej
02-21-2003, 08:18 PM
Originally posted by Brian C

Hey Chris.....tell Matt C. I said Hi!

Don't worry, we'll make sure he gets the message....btw, how you doing Brian?

Brian C
02-24-2003, 08:21 AM
Pretty good Deej, Thanks for asking

The usual butterflies wondering if the design is on the correct track or whether we "missed the boat" we'll find out in a few weeks.

How are you doing this year with your other project and getting along without FIRST as much? I remember you were torn in your decision making a few months back. Hope it all worked out for you.


Will we see you in New Hampshire?

Deej
02-24-2003, 10:05 AM
The project got put of till this summer....and I somehow managed to find some time to do some FIRST. I'll most definately be there in New Hampshire, and with some luck, Seattle and Houston. See ya soon!

Kristi
04-15-2003, 08:08 AM
Congrats on the Excelence in Control! Would you care to give us a clue how you did this... and how you stayed within the electronics budget?
- Kristi, Team 30

Deej
04-15-2003, 10:13 AM
We did a few things here.....the first is that we designed an Inertial Navigation System (INS for short) that used a gyro to detect our heading on the field. It also used a dual axis accelerometer (x and y axis). Based on knowing our lateral accelerations and conversions and our heading, we are able to know where exactly we are on the field at all times (well the computer knows). this was our main navagation device during autonomous mode. this system also allowed us to use "heading hold" which when utilized, maintained a straight track, and if we were pushed off course, the motors would automatically course correct. In Houston, our INS system was not working correctly (funny how in seattle the weekend before it did), but the judges were really impressed with our control setup for the operators. we were able to use 16 different preset autonomous modes, using a rotary switch, and also determine where on the field we would go using x and y coordinates on our box. we also had a few automatic functions, like restow our wings, extender, and pitch in one button push. the different autonomous modes also allowed us to see out on the field who we were up against and paired with, and make ont he fly decisions, like after we set up on the field before we stepped away from the controls, see where the opponents lined up, and make a quick change if necessary before the match. Hope that answers the questions.

WernerNYK
04-17-2003, 12:17 PM
Adding to what DJ said, there are a lot of little "hidden" control components that we had. We have 1 keyswitch on the robot that turns off the rotating light; the benefit of this is to save our battery while working in the pits, as well as just to get rid of the annoying flashing. And you dont have to worry about losing the breakers or forgetting to put them back in before a match. Even if we forget to take out the key and leave it in the disabled position, as soon as the RC is reset the light will come back on.

There is another keyswitch on our operator controls which enables "maintenance mode," allowing us to do such things as easily calibrating the joysticks.

We also built custom, intuitive controls to operate our wall. It consisted of two handles coming vertically up, with horizontal handlebars on each. To pitch the wall forward, just lean the assembly forward, to pitch backwards, lean back. To open the left wing, twist the left handle to the left; to open the right wing, twist the right handle to the right. To extend, click a toggle switch at the end of the right handle up, click it down to retract the extension.

And, our pivoting electronics board allowed for easy access to all electrical and mechanical components of the robot.

Frank(Aflak)
04-30-2003, 09:42 PM
what discipline would know how to build a custom circuit to process the signals from two accelerometers and one rotational acceleration sensor into an (x,y) position (or displacement from starting position) and a heading?

like would I want an electrical engineer, a computer engineer, a software engineer, or what? I would really love to see something like this on our robot next year. It its so . . elegant a solution to autonomous mode.

Dave Flowerday
04-30-2003, 10:26 PM
Originally posted by Frank(Aflak)
what discipline would know how to build a custom circuit to process the signals from two accelerometers and one rotational acceleration sensor into an (x,y) position (or displacement from starting position) and a heading?
Computer engineering is probably what you're looking for. Computer engineering is somewhat of a mixture between electrical engineering and software engineering. When I studied computer engineering in college (at the University of Michigan) we primarily learned about digital electronics and low level software, which is exactly what a microcontroller-based custom circuit is. You will also get exposed to a lot of computer science as well as a little bit of analog electrical engineering.

Also, the differences between electrical engineering, computer engineering, and computer science can vary depending on the college, so if you're thinking about majoring in any of those you should make sure you understand what that means at that particular school.

ColleenShaver
05-01-2003, 02:10 AM
The brains behind it on our team were mainly either computer science or electrical engineering (students and graduates alike, not all from WPI).

Dave's got good advice-- check out what it means for the school you're interested in attending. For instance, at WPI we have "Electrical and Computer Engineering" which would probably be the department that something like INS would fall under. You would need to check out course catalogs at each college to see what is required and available. At WPI, you could look at past MQPs (Major Qualifying Projects, known at many schools as senior projects) and their topics to get a feel about what people in that major study.

And if you're just looking for someone on your team that could build something like INS, once that portion of the team is over some of our major off-season activities (they write scoring software, etc for our off-season comps and others), they may be willing to put in some time explaining the basics of it.

Frank(Aflak)
05-06-2003, 07:47 PM
Dave, thanks for the help and advice. On the issue of CE vs EE, a little interesting bit of info is that Carnegie Mellon has a major EECE, or electrical and computer engineering.

heh. Now I need to go kidnap us some engineers. Or just ask them to help.