There has been some talk here before about the RTK-Lib project, which allows much greater GPS positional accuracy. Now someone has ported RTK-Lib the Android and it's in the Google PlayStore for free:
https://play.google.com/store/apps/details?id=ru0xdc.rtkgps
So apart from an Android device (or Android-x86 image) you need a raw capable GPS (Skytraq S1315F-RAW or something fancier) with decent antenna and a correction feed (or second GPS)... and volia cm level accuracy for your robotics projects.
Don't know for certain whether this can run headless once configured.
There's also a 'report GGA to base' function, which might allow the computation to be done in 'moving base' mode and then the solution's position relayed to the rover.... if you don't want to put extra processing on the drone.
Enjoy, Simon
Comments
Hi Will - I just had a look at the PA300 here : http://www.navtechgps.com/assets/1/7/hemisphere-gps-pa300-datasheet....
Looks awesome - full RTK solution at a 5x8cm footprint using various correction messages(RTCM, CMR+). Antenna included - and it has a very good initialization time for the antenna size. This is coming in at a 5 to 8 times less than the other brands - thanks for sharing. I just purchased a Trimble R4 receiver at around $10k. OK, it is housed with a battery solution, but still......I like the way this is going. At this rate it might become affordable to mount RTK systems on my plant.
There are a few L1/L2 boards out there now in the 1k-2k range now. I recently checked on the the PA300 miniEclipse Smart Antenna GNSS Module ans was priced at $1658.00 (L1L2, 1Hz, raw data). (has antenna built in)
A year or two ago the P200 miniEclipse II Module, L1 GPS, 1 Hz, Raw was $450, once you added L2 it added another $750. I'm not saying these are cheap, but sure a lot less than they were 5 yrs ago. These prices without RTK, but using rtklib.
L2 makes that RTK initialization very very fast, otherwise as Antonie noted, it takes a long time to get it fixed (unless you tell both receivers exactly where they are).
Sounds like you got a great deal with that Topcon receiver and I'm glad you're having fun with it.
Antonie had noted that the commercial offerings in this space can be expensive, but if someone is looking for a small dual frequency board there is the Topcon B110. I believe the starting price is ~$1000 for L1+Raw, increasing depending on what options you have enabled (L1/L2 RTK is ~$6000).
As I noted before here you can get quite nice results with L1 system, but note the antenna selection is quite important.
Simon
Then I start to config the receiver to have raw jps (Javad proprietary) output to read by RTKlib,by inputting GPS base from my local Ntrip server (Myrtknet) it got FIX position less than one minute. Then I set output NMEA to virtual commport and reread them with surveying software Carlson SurvCE...boom....Its ready to perform surveying, I test to perform topographical surveying for a stockpile, the result as good as firmware base RTK!!
Just for sharing
Yeah, that's where the catch comes in - to make this worth while you need receivers that can receive at least the L1 and L2 frequencies. Not only the receiver, but a decent antenna as well. Suddenly the price tag is running well beyond us hobbyist reach. The ublox 6T and 6P(I think) chips can be used to achieve a RTK fix, if you have a survey grade antenna attached and then it is only usable for static applications - only reads L1.
I think the world of RTK fix to the hobbyist is still way off, hey that does not stop us from trying. I had a 6T up on the roof in the week and achieved a RTK fix using RTKLIB within 20 odd minutes. I used a GIS L1 antenna. Luckily we have a network of bases in RSA - NTRIP casters, so getting a correction from a base is a matter of knowing the IP address...and its a free service.
I have not been able to find cheap (L1+L2+L5??) receivers out there and I've looked extensively. If you do find, please post here.
Thanks for sharing, Simon.
please can anybody suggest cheap pair of GPS receivers and a setup procedure to run this appication???
By taking raw measurements for each receiver, and passing them over the data link the system can compensate for the effects of the atmosphere. The position computed will be much more accurate but is referenced to the 'base' location, which is why you'd normally set the 'base' up over a known position (or survey it in)
For more information see: http://en.wikipedia.org/wiki/Real_Time_Kinematic
Cool.
But just to make sure I get what it does right:
I'm sorry for being such a noob, but I don't know the RTK-Lib project.