Affordable, Arduino-compatible, Centimeter-Level GPS Accuracy

Our NavSpark Indiegogo campaign brings latest GPS / GLONASS / Beidou / QZSS / SBAS satellite navigation technology and high-precision centimeter-level accuracy RTK technology accessible for everyone. Arduino programmable feature make it simple to use for electronic projects. Should be of interest for UAV applications.

If there is any question, I’ll be happy to answer here. Or can email us at Thank you!



Views: 3742

Comment by Greg Dronsky on January 16, 2014 at 8:43am

If this can work with APM i'm interested.

Comment by Liam Lambkin on January 16, 2014 at 9:30am

Same here, I have pledged anyways, It would be cool if we could install Arducopter on it, then it just leaves the sensors to buy :D

Comment by Ruwan on January 16, 2014 at 10:13am

Just curious, cm level accuracy comes only from RTK compatible NS-RAW? or from all of them?

Comment by Jeff Huang on January 16, 2014 at 10:15am

Since NavSpark has software development kit integrated into the provided Arduino IDE environment, users can easily program it to do other things at the same time GPS is working, or customize the NMEA output behavior to emulate other GPS module.

Different chipset vendor has different proprietary configuration commands. If the application does send proprietary configuration command, then there will be problem. But as APM is open-source, may be the GPS module configuration code section could be modified to accommodate...

Comment by Jeff Huang on January 16, 2014 at 10:25am

Only NS-RAW has carrier phase raw measurement output feature to work with RTKLIB to achieve centimeter-level accuracy. Offered as separate perk, to increase the chance of reaching funding goal. :-)

Comment by Rob_Lefebvre on January 16, 2014 at 11:48am

Looks interesting.  I wonder if inertial sensor fusion into the GPS fix, would help refine the position?

Currently we take whatever comes out of the GPS, and then fuse that to the interial data.  But if the GPS were to do it internally, it might help it actually get a better GPS position estimate?  

Comment by hotelzululima on January 16, 2014 at 11:53am

hmm I dont see any reason NUTTX could NOT run on it... ie sparc seems to be supported by nuttx..

so the next step is porting the HAL layer and the apm code itself.. the processor itself has more than enough memory and flash to do this... then you would need to control your servos might be best to see if the hardware is open design and simply modify their layout to have the standard APM peripherals or perhaps a daughter board(meh)...


Comment by Jeff Huang on January 16, 2014 at 4:25pm

Adding inertial sensor is helpful only in situation where there is signal blockage, providing continuous navigation for some time when GPS is not getting fix. Combining inertial sensor data with GPS measurement data to form Kalman Filter solution will give more accurate result than simply forming weighed sum of inertial sensor output and GPS NMEA output. Low cost MEMS accelerometers are noisy. For vehicle navigation, integrating accelerometer and gyro data to form relative movement, the solution quickly diverges compared to traditional gyro + odometer based relative movement estimation. The error for the former is in number of 100 meter error per minute; while for the latter is number of 100 meter error per 10 kilometer traveled, if not moving then no increasing error.


GPS gives absolution position. Inertial sensor gives relative movement, with expensive ones more accurate and diverge more slowly. For UAV application it’s under open sky, so won’t expect to have position accuracy improvement with MEMS inertial sensors.


Just Googled NUTTX, didn’t see Sparc-V8 support. Regarding porting of APM to NavSpark, note that there is only 1 PWM on Venus822 chip or NavSpark pin-out. If to drive UAV rotors, will need another low cost 8bit controller with enough PWMs and serial interface for control.


If the NavSpark Indiegogo campaign is successful, hardware details will be provided to allow users to maximize usage of the CPU and peripherals. Two sets of libraries will be released, one without GPS/GNSS, and one with GPS/GNSS. The library without GPS/GNSS support, users can have 100% use of all the hardware; information about the LEON3 Sparc-V8 processor used in NavSpark can be found at Anyone interested in porting to NUTTX, will need to start with this library without GPS/GNSS. The other library with GPS/GNSS, users can work within current structure without RTOS for the applications.


If someone really took the efforts ported NUTTX and APM, we’ll try porting GPS library to it and see if can run properly, release if without issue.


Except the GPS/GNSS library is released in library format (not in source code), the other parts will be opened to facilitate the adoption of SkyTraq's GPS/GNSS chipset and module. 

Comment by mP1 on January 16, 2014 at 7:17pm

How about a video where you walk with the reciever and have the gui next to it so we can see the cms increase slowly. Perhaps walking away and back so we can see it returns to the original value!

Comment by Jeff Huang on January 17, 2014 at 7:52am

The $25/pcs NS-RAW module on the NavSpark Indiegogo campaign enables user to easily setup and experiment with RTK technology using open-source RTKLIB on a PC or laptop. As the operation is not like using a Garmin handheld, the suggested video demo of in motion evaluation requires additional hardware setup, which we leave for users to experiment.


A paper describing in motion evaluation of low-cost RTK GPS receiver using RTKLIB by its author can be found here:


The setup needed to run RTKLIB is described in the user’s manual:


You need to be a member of DIY Drones to add comments!

Join DIY Drones

© 2014   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service