LN-270 / LN-251 GPS module with intergrated INS

[BADBOY13]

Hi,
Anyone heard of the LN-270 / LN-251 GPS module by Northrop Grumman?
http://www.nsd.es.northropgrumman.com/Automated/products/LN-270.html

It's a GPS module with intergrated INS(inertial navigation system).
I'm planning to use it on my robot.
If i'm not wrong the US army is using the LN-270.
Admin, do you know anything about it?
Like how much does it cost, is it compatible with the compactRIO / PXI (National Instruments) CPU that i'm planning to use?
Also, i don't understand the units used in the specs sheet, (i actually wanted to write a program and build a device that does similar calculations to achieve readings for positioning without the help of GPS until i finally found information on Inertial Navigation on the web and realized there are actually off the self module i can buy), what is CEP, PE and TTFF???

Can anyone help me with this?
Thanks alot

[JonHylands]

I hope you have a lot of money. I have no idea what it costs, but based on who makes it and what the market is, I would guess its in the $10-20 thousand range.

Here's an example of some open source-based products that do that:

http://www.rotomotion.com/prd_AHRS.html

You can buy a GPS module for about $60 now, and the parts for a six-axis IMU are about $175. Writing the software to integrate all that into a combined position estimate is a non-trivial thing, and that is most of what you are paying for.

- Jon

[Admin]

If i'm not wrong the US army is using the LN-270.
Admin, do you know anything about it?

I work for the Navy, with underwater robots that cant use GPS

I remember back when I first made this site I posted a link that lists robot parts the military uses. Cant seem to find it now . . .

[BADBOY13]

Thanks for the replies.

Jon:
money may not be that much of a problem as i'm still working my the budget and will most probable be sponsored.
hmm..
am i right to say that the INS module featured on the the link you gave me performs the same functions as the LN-270 by Northrop Grumman i mentioned?
However, surprisingly, based on the specs. given on the respective sites, the one you showed me seem to be much more accurate. Plus, according to the specs give, the system has NO Drift! Is that even possible?
Or is it because the accelerometer and gyro sensors are not very sensitive so they cannot detect small deviations?

Admin: ya, i understand that GPS cannot be used underwater. But you do you INS for navigation of your robots underwater right? What other ways do you use? Or you can't tell me because it's classified?
From your experience, how much does INS systems cost? Also, what about PINS?
hmm...
do you have any experience with LabView and compactRIO?
do you by any chance have any idea if the modules by Northrop Grumman or Rotomotion are compatible with compactRIO and LabView?

Thanks

[Admin]

ya, i understand that GPS cannot be used underwater. But you do you INS for navigation of your robots underwater right? What other ways do you use? Or you can't tell me because it's classified?

havent decided yet, its all a work in progress . . .

do you have any experience with LabView and compactRIO?

ive used LabView, but i wasnt very impressed with how it handled my specific requirements . . . i ended up building my own datalogger and writing my own software . . .
i dont know the answers to the other questions . . . sorry!

[JonHylands]

am i right to say that the INS module featured on the the link you gave me performs the same functions as the LN-270 by Northrop Grumman i mentioned?
However, surprisingly, based on the specs. given on the respective sites, the one you showed me seem to be much more accurate. Plus, according to the specs give, the system has NO Drift! Is that even possible?
Or is it because the accelerometer and gyro sensors are not very sensitive so they cannot detect small deviations?

They say "no drift" because they have implemented a very accurate Kalman filter that compensates for it. With a GPS providing a steady frame of reference, I'm sure they get virtually no drift. However, if the GPS signal goes away (like the robot has to enter a long tunnel, or is indoors), eventually it will start to accumulate.

Personally, I would go with the open source one - you're a lot more likely to get support without having to pay through the nose. I'm sure the LN-270 is better built, from better components, and certainly to milspec, and I'm equally sure the open source one isn't. But if it was my project, unless the funders are insisting on military grade equipment, I would save the money here and spend it on other cool toys, like LIDAR or one of the high-end stereo vision systems...

- Jon

[hgordon]

Jon -

If you were building a 6-axis IMU + GPS that was a nice tradeoff between accuracy and cost,  what components would you select ?  Also, is a 3-axis magnetometer built into the GPS, or is that separate component ?

Howard

[JonHylands]

I would use three ADXRS401s, and an ADXL330. I would also add in one of the Honeywell 3-axis magnetometers, although I'm not sure off the top of my head which one. The reason for the ADXRS401 instead of the 2-axis one is the Analog parts have internal temp. sensors on the die, which allows a good Kalman filter to compensate for temp changes, which is one of the major causes of drift.

For a GPS, look at Sparkfun - they have some awesome ones for under $100, that are tiny. Pick one of the new 20-channel modules, or even the new 32 channel one that has a 5 Hz update rate:

http://www.sparkfun.com/commerce/product_info.php?products_id=8266

If you're really trying to cram things into a small space, I would look at this:

http://www.sparkfun.com/commerce/product_info.php?products_id=8291

For a micro-controller, I would look at something like one of the ARM7 chips running at 60 MIPs, and use a couple 16 bit 8-channel serial (SPI) A/D converters.

You'd probably want to have this mounted inside an enclosure - you could probably get away with 1" high and 2" x 2" square, if you're careful.

- Jon

[JonHylands]

Gordon,

I don't know if you've seen it, but I'm in the last stages of preparing for production a much lower-end 6-axis IMU:

http://www.bioloid.info/tiki/tiki-index.php?page=6-Axis+Bus+IMU

This one is specialized for the Bioloid, but that's just software for the most part.

I'm going to sell the IMU board, populated, without the sensors, for $50. The sensor boards are basically two breakout boards from Sparkfun - the 5-axis IMU board, and a one-axis gyro board.

Its not nearly as accurate as what I described above, but its less than one cubic inch (final size is 1" high, 1.15" long, 0.75" wide), and a whole lot less expensive.

http://www.huv.com/blog/2007/04/imu-mounted.html

- Jon

[hgordon]

Thanks Jon -

That's more or less what I had in mind.  I was actually thinking about using an LPC2101 ARM7 for this type of design, though the A/D's are only 10-bit.  Are the sensors sufficiently precise that you need higher resolution A/D ?

I'll follow up with you about this via email.

Howard

[JonHylands]

Yes, if you want to do reasonably accurate navigation, you should use at least 12 bits on your A/D. There are people out there using the built-in 10 bit A/D converters (including me), and with the appropriate filtering, they seem to work. Like most things, more accuracy is better, but it really depends on your other constraints.

- Jon