Ardusimple simpleRTK2Blite heading and attitude post processing

Q&A forumCategory: QuestionsArdusimple simpleRTK2Blite heading and attitude post processing
bakalolo asked 4 weeks ago

We want to use the simpleRTK2B with the simpleRTK2Blite to get both position and orientation by saving raw data to a Raspberry Pi.  We have already tested a Pi with a single simpleRTK2B to save raw data and that works.  The raw data was processed using RTKLIB.  What would be the procedure to calculate pitch roll yaw (attitude) for post processing?
Should we simply calculate the position of the 2 antennas and subtract their Lat Long Height to get heading (attitude).

Ardusimple Staff replied 4 weeks ago

Hi bakalolo,
When mounting simplertk2blite on top of simplertk2b, the simplertk2b will output already RTK position and heading.
If you want to postprocess the data, it’s better to buy 2 simplertk2b and read raw data from both units, simplertk2blite doesn’t help in this case.

clive1 replied 3 weeks ago

>> subtract their Lat Long Height to get heading
Most things would generate a relative displacement in a NED format (North, East, Down, in metres). The ZED-F9P of the secondary would report this real-time via UBX-NAV-RELPOSNED

1 Answers
bakalolo answered 2 weeks ago

So if I understand correctly, the advantage of having the simpleRTK2 and the simpleRTK2lite coupled together is to get the heading info added to the  UBX-NAV-RELPOSNED messages. But I could also get heading by calculating the diff in the position of two independent simpleRTK2 systems. 
Is there a diff in accuracy between the two methods? Our tests suggest ~16 mm horizontal drift and ~100 mm vertical drift (after corrections). If the two antennae are about 1 m (1000mm) apart, that drift corresponds an angle drift of 16/1000 = .016 rad =  0.92 deg. Does this heading calculation seem about right to you?

Ardusimple Staff replied 2 weeks ago

About first point:
-If you have a base station with a short baseline, the performance will be very similar yes. In a long baseline conditions, the moving base will give you more accurate results.
-If you don’t have a base station, then it’s different. the moving baseline feature will give you still accurate heading.

About the second point, your calculation of accuracy is quite accurate. In the zed-f9p datasheet there’s a nice graph showing antenna distance vs heading accuracy: