A reader wrote me about occupation times for RTK work, and it’s spurred a conversation I think will be interesting to you and perhaps a little controversial. It seems that most GPS/GNSS users have developed their own opinion based on their own experience.
The discussion has several points, but the one I’d like to address in this column is the occupation time for RTK points. I’m not referring to the topo type of point where you’re collecting somewhere between 1 to 5 seconds (and averaging) of data, but rather the RTK shot where you want the highest confidence and accuracy in the RTK position.
I realize that most, if not all, manufacturers advise (or design into their software) that 180 seconds of data is sufficient for an RTK shot where the purpose of that point is to establish secondary control.
The reader offered that he “couldn’t imagine that we are getting a good solution with anything less than 120 epochs.”
I scratched my head on this one, and even checked with a few GNSS engineer friends of mine about the upside of occupying a point with RTK for 180 seconds (assuming a 1 Hz rate) rather than 30 seconds, or even 15 seconds for that matter.
First of all, there are several assumptions in this conversation:
- You have clear view of the sky.
- It is clear of multipath-enabling obstructions.
- Six or more GPS satellites are being tracked with a low PDOP.
The first thought in support of 180-second occupation time would be multipath detection/mitigation. Of course, some multipath isn’t going to be detected, but if it is, it’s going to happen in the first few minutes. However, if you’re really concerned about accuracy, you wouldn’t be using GPS to set control in a GPS-unfriendly or marginal environment in the first place!
In lieu of a 180-second occupation time, I see greater upside in occupying 15-30 seconds twice during the day at time where the GPS constellation is significantly different, but still with six or more GPS satellites with a low PDOP. This would do more for my confidence in the accuracy of the position than one session of 180 seconds.
Also, there’s discussion of a 180-second session taking five minutes to collect because it rejects measurements that exceed the tolerances set in the receiver. I don’t like this idea. It tells me measurement isn’t stable enough to begin with (unless you have the receiver set to some extraordinarily low tolerance). I’d rather set up over the point, let the RMS values stabilize (should be just a few seconds), and then record a 15-30 second shot.
Of course, we are only talking about another 2.5 minutes of occupation time, and I’m guessing that most wouldn’t mind spending that on a point designated for secondary control. However, I do see that as the economy continues to put pressure on companies to keep the costs down, that pressure will be put on the field crews to look for time savings. I think occupation times will be one area and not just on establishing control points, but when collecting topo, too.
I’ll continue on this subject in the next newsletter after discussing more with my colleagues and hopefully hearing comments from you. Also, it’s worthwhile reading a draft document published by the National Geodetic Survey outlining that agency’s guidelines for single baseline RTK users. It discusses, among many other guidelines, the issue of RTK point occupation mentioned above. You can view or download it here.
Leica/NovAtel Follow-up on RTK Occupation Times
Following up on my last newsletter, a few folks wrote me about my comment in the Sokkia/Topcon discussion where I noted that Novatel was now owned by Leica and how it would impact the Sokkia/Novatel joint venture named Point, Inc. Several readers pointed out that Leica doesn’t own Novatel, but rather both companies are owned by Hexagon AB of Sweden.
I understand the technical aspect of one company “owning” another and I certainly misstated that. I was writing more from a strategic view.
One reader commented that “Both Leica and NovAtel are part of the same group, but they do business at arm’s length, as you would call it. NovAtel supplies Leica with core technology in a standard supplier-buyer relationship.”
I think it is a little cozier than that. I don’t believe that Hexagon would have touched NovAtel if they didn’t own Leica already, and I think that Leica folks had a lot to do with encouraging the acquisition and were probably intimately involved throughout the due diligence process.
But I think a good point is made that NovAtel is still committed/focused on being an OEM supplier of precision GNSS receivers. While being owned by the same parent company as Leica hasn’t helped their image as an OEM (original equipment manufacturer) of GNSS receivers for precision (survey and other markets), they are still very active in that business and seem committed.
A couple of other notes and I’m done with this subject for the time being:
- If you recall, Spectra Precision is owned by Trimble. It’s not surprising that the Spectra Precision Epoch 25 RTK system was designed using a Trimble GPS receiver. However, the new Spectra Precision Epoch 35 RTK system announced in January uses a NovAtel GNSS receiver. Quite uncharacteristic of Trimble and maybe even unprecedented in their high-precision business (in the past, they’ve used some other GPS receivers in their low-precision GPS products).
- In other significant NovAtel news, NovAtel announced last week that CEO Jon Ladd is leaving NovaAtel and taking a “strategic advisory role” with Hexagon. Personally, I have a lot of respect for Ladd. After NovAtel suffered through years of finance and administrative type CEOs that floundered, he’s a true GNSS guy and was the right person for the job. He’d been CEO at NovAtel for seven years. Prior to that, he was a key technical executive at Ashtech. Ladd is being replaced by Michael Ritter, who most recently was an executive in Trimble’s Engineering and Construction group.