Society of Robots - Robot Forum

General Misc => Misc => Topic started by: h3ro on September 14, 2007, 04:22:04 PM

Title: Accuracy of GPS?
Post by: h3ro on September 14, 2007, 04:22:04 PM
I was just wondering what is the Accuracy of GPS system?

Is it possible to get them down to 0.5 meters? And if its possible, what price range are we looking at?

Thanks:)
Title: Re: Accuracy of GPS?
Post by: Rebelgium on September 14, 2007, 04:31:40 PM
Interesting question, I'd like to know to. :)
One additional question :
Does that accuracy stay when you're indoors? And do you even have a GPS signal indoors?
Because my GPS for my car doesn't receive anything indoors...
Title: Re: Accuracy of GPS?
Post by: Admin on September 14, 2007, 08:35:36 PM
As of this writing, the best commercial GPS gives an accuracy of about 5 meters.

Its a fast changing technology though, so prices and unit sizes are dropping really fast. I'd expect accuracy to improve by 2 or 3 more meters in the next decade or two.

If you need really high accuracy, you have $$$$, and its an indoor robot, consider this:
http://www.acroname.com/robotics/parts/R275-NS-DEV-BUNDLE.html
Title: Re: Accuracy of GPS?
Post by: JonHylands on September 15, 2007, 05:31:20 AM
If you have good WAAS coverage where you are, you can get about 1m with a standard ($150) GPS that supports it. I have a Garmin Geko 201 with WAAS that gets about 1m in my backyard.

If you have a lot of money to spend ($10,000), you can get 10cm accuracy with DGPS.

- Jon
Title: Re: Accuracy of GPS?
Post by: Rebelgium on September 15, 2007, 06:41:41 AM
In my backyard I also get about 1m with my commercial car GPS, but for now GPS is waaay to expensive for me ;)
Title: Re: Accuracy of GPS?
Post by: Steve Joblin on September 15, 2007, 08:04:15 AM
A buddy of mine works at a place that makes GPS "analyzers"... they analyze the amount of potential error based on your location and altitude.  For example, where I sit, a GPS reading may be +/- 1 meter in accuracy... where you sit, the same GPS may only be +/- 5 meters in accuracy.  It all depends on where you are as to how accurate of a reading you get.
Title: Re: Accuracy of GPS?
Post by: Admin on September 15, 2007, 09:29:18 AM
JonHylands, I was repeating what I was told a week ago at this presentation given by a GPS expert at the navy base I work at . . .

So I looked up the specs on the Garmin Geko 201 manual
http://www.garmin.com/manuals/Geko201_OwnersManual.pdf

GPS Accuracy: <15 meters (49 ft) RMS
DGPS (USCG) Accuracy: 1-5 meters (3-15 ft) with DGPS corrections
DGPS (WAAS) Accuracy: 3 meters (10ft) 95% typical with DGPS corrections

So as Steve Joblin was hinting at, you got really lucky on your accuracy. But still, this is a better accuracy than I was led to believe . . .

Anyway, the expert was telling us that WAAS and DGPS technology will be greatly improved in the next decade or so. He also said there was new technology being worked on that factors in the atmosphere conditions too (which affects speed of light/GPS transmissions).
Title: Re: Accuracy of GPS?
Post by: h3ro on September 15, 2007, 09:32:36 AM
So using a GPS for navigation on a small indoor robot would not work very well then I guess. 1-5 meter would be the make you wonder if its even in the right room :P

Will the offset be constant? Say if you do a reading an the GPS tells you that where you are, but in reality you are 1.3 meters longer to the right. If you move and do a new reading, will you still be 1.3 meters from the new position? Or does it change all the time?
Title: Re: Accuracy of GPS?
Post by: JesseWelling on September 15, 2007, 09:56:41 AM
From what I've been studying it the average position changes slowly over time...
So a reading you took yesterday at one spot might be 5m off the next day.
It seems like there are two kinds of problems you have to deal with:

1. Day to day drift, which can be fixed with an new offset from a known position on power up.

2. Noise coming in (this is the biggest part of the error) so that even if you are sitting still
    your GPS signal might tell you you are moving around a bit.

I've been studying (among other things  ;) ) way of improving gps and combining it with IMU/Odometry,
and it seems workable but heavy on the maths (for me), but where is the fun if there is no challenge right?

Any ways here are two of the most suitable GPS recievers for mobile robotics that I've found:
U-Blox demo kit (http://www.abacuscity.ch/abashop?i=rUkGH1CGKpM4LVrIaXR2&s=142&p=productdetail&sku=71)
ETek GPS from SparkFun (http://www.sparkfun.com/commerce/product_info.php?products_id=8266)
Title: Re: Accuracy of GPS?
Post by: Admin on September 15, 2007, 12:38:52 PM
Quote
Will the offset be constant? Say if you do a reading an the GPS tells you that where you are, but in reality you are 1.3 meters longer to the right. If you move and do a new reading, will you still be 1.3 meters from the new position? Or does it change all the time?
Quote
From what I've been studying it the average position changes slowly over time...
So a reading you took yesterday at one spot might be 5m off the next day.
The satellites are always moving, and your gps would get a lock on different satellites at different periods of time. Atmosphere affects readings, which also changes. Nothing to keep the error constant . . .

Quote
So using a GPS for navigation on a small indoor robot would not work very well then I guess. 1-5 meter would be the make you wonder if its even in the right room
yeap. GPS also has a slow update frequency. the cheap ones update once every second or two - your robot wont know its run into something till its too late . . . which brings me to jesse's point:

Quote
I've been studying (among other things   ) way of improving gps and combining it with IMU/Odometry,
and it seems workable but heavy on the maths (for me), but where is the fun if there is no challenge right?
just so you know, that expert said this is the future of gps. gps and IMU's compliment each other well, it only requires kalman filters to make it happen . . . still too much math for me to fully understand, tho . . .
Title: Re: Accuracy of GPS?
Post by: JesseWelling on September 15, 2007, 02:45:53 PM
Yea Kalman filters are crazy full of weird math that software engineers hardly even touched in school.
Good thing I work at a place that's full of control theory geeks...  :P