GPS's, Is that it?

Go To Last Post
22 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I was think about get a GPS module for my robot project but I was wondering if GPS modules in-general only give coordinates, or can they give directions to way points? Because if they only give the coordinates then it sound like there would be a lot of programming to get way point functionality, too much for me.

Life Is Like A Bucket Of Chicken.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Depends what you mean by directions, GPS does NOT give you a bearings(NESW), unless the receiver is moving(I mention this because it seems to be a common misconception).

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

And they hardly can gave you anything if inside a building (just in case you plan to use this robot inside some). Precission is not the best, and they usually blend their data with IMU's (inertial measurement units) using something like Kalman filters, etc. That is, damn simple programming is awaiting for you if you want precission.

Guillem.
"Common sense is the least common of the senses" Anonymous.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

GPS *can* give you a "bearing" to a waypoint, whether it is moving or not, but can only give you a "heading" (and speed) if it is moving. Coordinates are continuously available.

Tom Pappano
Tulsa, Oklahoma

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

tpappano wrote:
GPS *can* give you a "bearing" to a waypoint, whether it is moving or not, but can only give you a "heading" (and speed) if it is moving. Coordinates are continuously available.

OK. What I meant was that a stationary GPS unit has no idea which way is north.

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thinking about it, John, it *knows* which way north is - it has an ephemerides, it knows which satellites are in which direction... it just doesn't know which direction to point! :mrgreen:

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

barnacle wrote:
Thinking about it, John, it *knows* which way north is - it has an ephemerides, it knows which satellites are in which direction... it just doesn't know which direction to point! :mrgreen:

I'm sorry, you've completely lost me there...
But then I don't have GPS.

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Imagine you had only one eye, with a complete 360 degree field of vision--and you lacked proprioception.

How would you know which way your legs are pointing without taking a step?

Elect reputable scientists!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Senacharim wrote:
Imagine you had only one eye, with a complete 360 degree field of vision--and you lacked proprioception.

How would you know which way your legs are pointing without taking a step?


To whom are you addressing this question?

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My point, John, is that it knows which satellites are visible and where they are just from knowing the time - that's how it knows *where* it is.

Because it knows the position of the satellites, it knows where north is, in a theoretical sort of way. But it doesn't know which way it's pointing, because it has no external reference to its immediate orientation.

If it moves, it knows which way it moved, but it doesn't know whether it moved left, right, forwards, or backwards - it doesn't know whether it's upside down or not.

So it knows where north is in the same way I know it's in the direction of Polaris... when I can't *see* Polaris. But like me, it can't stick an arm out and point north.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello Taco,

A typical civilian GPS will give you about 10 meters accuracy in your location. Taking several readings and averaging them will give you better accuracy, but at the expense of time. Small hand held and vehicle GPSs that plot your course on a road have additional info, the road data base, to help them know where you are, (presumably on the road, not driving down the median, or the shoulder...).

For mega bucks you can get down to 1-2 cm accuracy. But you can't buy one of those units at Spark Fun Electronics...

If you chose to use GPS then check out this site: Ed Williams Aviation Formulary V1.44 . This has a GREAT collection of information for calculating distances and headings and chords and simplistic approximations, and, and, and, ... in the spherical domain.

I ran a bunch of algorithms on a PC before I moved to a microcontroller as it was much easier for me to feed it known input data and validate the results on a PC.

Good luck with your project!

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

barnacle wrote:
My point, John, is that it knows which satellites are visible and where they are just from knowing the time - that's how it knows *where* it is.

Because it knows the position of the satellites, it knows where north is, in a theoretical sort of way. But it doesn't know which way it's pointing, because it has no external reference to its immediate orientation.

If it moves, it knows which way it moved, but it doesn't know whether it moved left, right, forwards, or backwards - it doesn't know whether it's upside down or not.

So it knows where north is in the same way I know it's in the direction of Polaris... when I can't *see* Polaris. But like me, it can't stick an arm out and point north.


Yes. I already knew all that.
My point was that a stationary GPS unit will not replace a compass.
I may have expressed myself badly.

Four legs good, two legs bad, three legs stable.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think I will also add a compass.

This is the GPS I was thinking of getting if it has the aforementioned functionalities, like way point bearing, and direction of GSP if it know which direction you are moving?: http://www.weirdstuff.com/cgi-bin/category.cgi?item=20164#

Life Is Like A Bucket Of Chicken.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

John_A_Brown wrote:

I may have expressed myself badly.

Not at all, John; I was merely amusing myself with the philosophy of the device.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

A typical civilian GPS will give you about 10 meters accuracy in your location.

Surely that's an absolute accuracy against a map datum such as WGS84, rather than a relative accuracy? Or are you suggesting that from one reading to the next in a fixed location the position might suddenly jump 10 meters? If this is all about navigating a robot through a maze then I'd have thought the starting point in the maze would be the datum and further readings would be relative to this (not WGS84)?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Or are you suggesting that from one reading to the next in a fixed location the position might suddenly jump 10 meters?

It can jump around or it can drift. You can help by using differential techniques. Have a second gps in a fixed position and calculate the moving gps's position by observing the difference from the fixed unit. Navigating a robot over a small course may not be easy due to the limits of precision available in the messages themselves. My gps gives me thousandths of a minute.

Tom Pappano
Tulsa, Oklahoma

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Does anyone know if the GPS module I mentioned has the the 'way point bearing' function?

Life Is Like A Bucket Of Chicken.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As the others have mentioned - most likely not.

Cliff - As Tom mentions, it can drift or jump. The Novatel receivers I use have a fancy utility that draws a snail trail on a little grid. Over time you can see the readings group around the actual point. When satellites come into and out of view you can get jumps. The more you get into GPS, the more you find it isn't quite as magic as you'd hoped. When it is all said and done, the position it gives is an estimation. The more information it can use, the better the estimation is. The high end receivers use the L2 frequency - although this is military and commercial receivers can't decode it they can use the carrier for more information. With Glonass and Galileo satellites becoming available, these too can be used for more information if your receiver can cope with them.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I agree with Kartman's comments.

There are many PC GPS programs that will decode the data, show it on the screen, and show a plot of where you are, Sat S/N ratios and positions, etc. Pull one up and watch your stationary position bounce all over the place!

One thing to note is that the manufacturer(s) sometimes "filter" the output data to smooth out the "Jumpiness" of the position/speed/etc.

On some modules you can adjust this parameter, and get the actual readings every second, (or whatever the update rate is for that unit), with no filtering.

You can also adjust how far "above" the horizon a satallite must be in order to be eligible to be used in calculations.

Back to Taco_Bell,

The Rockwell unit you are looking at is VERY OLD technology. Newer chipsets, front end signal amplifiers, and signal processing algorithms will blow this unit out of the water. I've never tried to guide a robot with GPS, but if I were to do so I would pick a SiRFIII module for starters, (Not trying to start a GPS war here, but making a suggestion).

I did not see a data sheet for the Rockwell unit provided on the site, and I did not google for it, but you can bet, (and you can check), that it will output an RMC data packet, (plus some others...). This "Recommended Minimum Navigation Information" will include the following:

    Field Number:
    1) UTC Time
    2) Status, V=Navigation receiver warning A=Valid
    3) Latitude
    4) N or S
    5) Longitude
    6) E or W
    7) Speed over ground, knots
    8) Track made good, degrees true
    9) Date, ddmmyy
    10) Magnetic Variation, degrees
    11) E or W
    12) FAA mode indicator (NMEA 2.3 and later)
    13) Checksum

So yes, you get the direction the GPS is moving, Degrees Magnetic. This is based on the GPS actually moving. As others have said, if it is sitting still, a GPS is NOT a compass, and does not know which way the robot is pointing. Head your robot full speed due North and the GPS will tell you this.

I have a list of GPS data Packets on my web site, and you can google for this list as well. GPS Data Packet Descriptions .

Be sure to check the data sheet because different units put out different data packets. All of the units I've ever worked with put out the RMC and others.

There is a lot of redundancy between the packets.

Note, also, that this "Great Price" GPS does not include an antenna. Be sure to look for an antenna and include that in your overall GPS price. Make sure you select either a powered antenna, (built-in LNA), or not, as required. If it has a built-in LNA is it 5V or 3V powered? Does the GPS board supply the power to the antenna connector, or do you have to power the antenna separately? What kind of connector is use, there is not a "Standard", many different types are used.

I would think, (my oppinion, again, without any personal experience), that GPS alone would be very poor for navigating a robot about a small, indoor course.

Integrated with other sensors it would work well for navigating across Km/Mi(s).

That all said, I still believe GPS is truely one of the Great technological marvels of our time, on many levels of analysis!

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Spark Fun Electronics, SFE GPS's , have a bunch of GPS modules which are available in single quantities, and have their data sheets listed, also.

Some break out boards are also available.

I would not try to design my own board with the LNA chip and chip antenna unless I had a VERY good reason to do so.

There is a GPS section on the SFE's Forum.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:
Note, also, that this "Great Price" GPS does not include an antenna. Be sure to look for an antenna and include that in your overall GPS price. Make sure you select either a powered antenna, (built-in LNA), or not, as required. If it has a built-in LNA is it 5V or 3V powered? Does the GPS board supply the power to the antenna connector, or do you have to power the antenna separately? What kind of connector is use, there is not a "Standard", many different types are used.

I would think, (my oppinion, again, without any personal experience), that GPS alone would be very poor for navigating a robot about a small, indoor course.

Integrated with other sensors it would work well for navigating across Km/Mi(s).

That all said, I still believe GPS is truely one of the Great technological marvels of our time, on many levels of analysis!

JC

Is an antenna absolutely need, so can I test the GPS before investing more into it, remember I have never used a GSP module before?

Life Is Like A Bucket Of Chicken.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Is an antenna absolutely need, so can I test the GPS before investing more into it, remember I have never used a GSP module before?

Yes, you need an antenna.

Without one it would be rather like testing a new car, without the engine installed...

You need a properly selected antenna to pick up the very weak GPS signal to feed the GPS receiver module. Without one you will never get a signal from the satellites and the module will never give you a position report, dtae, time, etc. It will, (usually, anyways), continue to output data packets where all of the fields are null.

This is one reason why a number of the GPS modules from SFE are so convenient, they include the antenna.

JC