I’d like to deploy a remote robotics system. I already have two F9P receivers (both mounted on the robot, providing yaw), but I plan on moving away from any nearby base stations. I might also need to jump around. The only thing I can reliably say is that there will be cellular service.
The options I’ve been looking at are: a.) streaming RTCM data via cellular and hoping that ~15km isn’t too far out for a fix and b.) setting up a base station with an XBee module to survey in and then transmit corrections directly. With respect to those options, I’m concerned that cellular (4G/LTE) may not be enough, and I’m also concerned that the XBee may run into issues with transmission rates (RTCM requires 4kbps, which is more than LoRA can provide; I’ve also never used XBee).
All that being said, I’m considering the combo board. I really like the idea of being able to show up anywhere and have it work immediately. That being the case, I have some questions:
Do correction services provide great results? I don’t know where their base stations are, so I’m not sure what performance I’d be buying into.
How would they compare to, say, the XBee option?
In terms of reliability, is this the no-brainer option?
I already have the two antennas that come with the RTK-SMA kit, but it seems like they wouldn’t work for the NEO-D9S. I’d prefer to not have to buy two antennas; if I buy one for the moving base / combo board, would I need another for the rover?
Lastly, the most important question is: it seems like corrections via L-Band are 0.2 hz. Are the corrections even usable for robotics?
After buying a suitable antenna, cable, mount, and the combo board, the cost is rather significant for the NEO-D9S. Is it worth it? Am I missing anything?
Well a lot of things to consider.
Robustness comes from options, in depth…
You can get SPARTN/PointPerfect service via Internet/Cellular, you can test that.
The ANN-MB isn’t an L-Band appropriate antenna, where as the ANN-MB2 is. There are also a host of Beitian antenna supporting all GNSS bands and L-Band.
The L-Band satellite is low to the southern horizon within the US, better honestly to have a static, directed antenna, to pick this up. It’s sending a Model for the satellite and atmospheric performance at 2400 baud. Not quite the same as local observational data via RTCM3, but probably closer to 5-6 cm.
Would be good for establishing a localized base station. You’re now averaging several cm rather than several metres.
LoRa radio has bandwidths from around 1 kbps to 22 kbps, LoS being frankly one of the most critical aspect.
With the Combo Board, you could evaluate, use to establish a fixed base for your Rover(s) quickly, furnish RTCM3 and SPARTN to them. Work on your radio comms/telemetry, potentially acting as a cellular gateway to them.
What a great response! I think I’ll hold off for now on the L-Band. 5 seconds between corrections is simply too much time for me to feel comfortable relying on it. I do very much like your idea of using it to set up a fixed base, so I’ll probably try to do that at some point.
5 seconds between corrections is simply too much time for me to feel comfortable relying on it.
That’s not really how carrier measurement / constraints work. On the moving-base front it’s the fact the base is moving which is the data with the highest criticality.
The RTCM3 MSM data are observations, not technically corrections.
The SPARTN data is primarily a model constructed from observations in signals from dozens of sites meshed over the coverage area(s). The L-Band cycles with data from the grids per the GAD, the dial-up method can constrain to only those covering your current location. In any case the longevity of the data covers the bandwidth concerns.
Provided there’s no loss of carrier measurements an RTK solution can be pushed out to a minute or more. Degrading to SBAS is the most disruptive.
Wow, you completely eliminated a misconception I had. I think I was confused because in order to get orientation measurements the moving base needs to broadcast corrections at the same rate as the rover, but I guess a 1Hz could by paired with a 5Hz rover (or moving base/rover combo). This article also helped, in case anyone is interested:
Regardless, thanks so much for your great responses. I’m a little tight on budget, so I think I’m going to try going down the fixed base / radio pathway.