Hi. I have a Piksi right now, thinking of changing over to the Reach.
Couple questions. Can it talk with a APM2.6 to navigate a Rover? Mainly does it have standard NMEA output?
I am in NC, there is no chance of seeing the horizon with hills, trees, and buildings in the way. How well will the reach work in this environment? I can usually get about 7-8 satellites.
Can any serial connection be used to link the two radios or is WiFI the only method?
Thanks,
Charlie
For the first question, I have no clue, the second, yes, for me at least, I can keep a float for a good portion of the time as long as the sky is clear. And yes different connections can be made between the base and rover including serial, refer to http://docs.emlid.com/reach/hardware-integration/
Hi Charlie,
We are working on integration with APM-based autopilots, it should be in a beta stage in a couple of weeks.
Reach already outputs NMEA, but NMEA does not provide enough data for APM, so we decided to add a binary protocol to provide better integration. More information on this topic is available in another thread.
Reach can work in such environments, but it really depends on how satellites are visible in this area. If they are high then some trees and small buildings should be okay.
Ultimately I will not be using the APM and putting in our own processor, just starting with the APM since its so easy to get up running with the Planner, etc.
Along the lines of accuracy, How many satellites is enough? So If I get 7 satellites is that enough. What kind of accuracy should I expect. Does the GPS drop or weight differently signals that are very close to the horizon and have a lot of atmosphere shift, etc. I am trying to get this going for a mobile robot that can be inserted into many different outdoor situations/locations and not require a PhD thesis to set up. Each GPS will not always have the same view, especially in a urban environment, going behind buildings, etc. I plan on using other sensor information, but GPS is the primary one. So if I read the number of visible satellites with the robot sees and it drops to 5 or 54, etc. would I know the approximate inaccuracy?
Thanks,
Charlie
