No I can definitely see the benefit. Our eBee has done very well for us as most of our projects are in the 200-400 acre range. I’ll probably have to give the eBee plus another try, but I still don’t think its possible to get consistent results with out GCPs. It takes many flights and hours of processing to get comfortable with the results.
I’m interested in the quad copter as an additional tool because of the new FAA regulations allowing flights 200’-AGL in previously restricted areas. I can also control the flight path in residential areas where I can restrict the flight to just over the streets and not people’s back yards.
This is a great thread. We’ve come full circle in the last year. For the longest time we were hell-bent on RTK. Our standard set up is CORs correction pushed to the rover (L1/L2 GNSS) via telemetry. This works great when you can connect to CORs and your telemetry link never drops. We hardly ever experience these conditions for 100% of the flight time in our hilly topography. So we are moving toward a PPK solution. Fortunately, we have a survey grade base that can always be positioned where it can be fixed with CORs. The other advantage of this set up is it minimizes latency. The GEO tagging process is more time consuming but worth the effort.
Even with PPK or RTK EXIF data, we will continue to use GCPs for the foreseeable future. We need to validate our airborne data and results with widely accepted surveying practices, which can only be done with GCPs. Our surveyor will not sign his name to a project without GCPs.
I am bit befuddled by the DJI 210 conversation…Pix4d will process any image with EXIF data, regardless of the source. Now the accuracy of the EXIF data is an entirely different subject and perhaps that was the intent of the original question.
1 Like
Our company just purchased the Matrice 210 RTK, and we would be really disappointed if Pix4Dcapture didn’t eventually support it. We rely on the software to do all sorts of mapping and modeling.
I spoke with a DJI rep back in July and he told me that the 210 is essentially the Inspire 2 in a rugged airframe. I am not a programmer, but I would assume that should make it an easy adaptation.
Hi Dave - thats a bold step and an expensive investment. Whats your opinion on the whole RTK adaptation by DJI? Most of us are used to RTK being used in a conventional survey environment (base station over a known point or benchmark, rover etc). My understanding from what I have read is that DJI are not intending to use the RTK features in a conventional sense for survey, but as a way to control the position of the aircraft in the air with respect to drift and close proximity flying to structures.
Would be good to hear your impressions and how you see this product being used. Maybe you could explain why your firm elected to invest in the RTK model rather than the single camera 200 non RTK model?
Hi Dave.
I’m very interested in DJI 210 RTK features: is it really useful? Than RTK features are intended for survey: in other words can it substitute RTK classic instruments in a conventional survey?
Hi Antonio
My understanding is that its NOT a survey grade GNSS RTK system in the way that you are expecting. Meaning that there is no survey tripod, tribrach etc and/or facility to set it up over a benchmark. Also no option for L1/L2 signal logging at both aircraft and base station.
RTK - the technique - is used for aircraft stability in the air. So instead of having sub meter drift vertically/horizontally, you have centimetric drift.
1 Like
Hi Philip,
I suspect you are right: the RTK feature is intended for “precision movements” not for geographic data acquisition.
1 Like
The confusion is caused by DJI using the term RTK which as we all know is just a method (Real Time Kinematic). As the survey industry has been using RTK and PPK for many years, when we see these letters we immediately assume it must obviously be a survey grade system.
As I have mentioned, DJI are using the technique of reducing drift of the aircraft by having it act as a rover from their base station. I can’t find any documentation which shows its a survey grade system or even will be.
One easy way to check this is next time you process some data, have a look on the GNSS antenna database to see if its listed as you would need to have the phase center parameters of the antenna. Also theres no indication of the ARP on the DJI base either - so you would not be able to accurately measure antenna height.
The most up to date list of GPS/GNSS receivers can be found here: http://gpsworld.com/wp-content/uploads/2012/10/GPS0117_S001-S023_RxSurvey2017R_SPREADS.pdf
1 Like
@Dave Schilke,
I would’t be too concerned if Pix4d doesn’t support the Matrix 210. There are plenty of other app makers, including DJI, that will. I haven’t used Pix4d capture in quite some time, but we quickly migrated toward a more feature rich app. We’ve been really happy with mapsmadeeasy’s application. We’ve been using it for over two years. I am not sure if the app current supports the 210 but I am sure it will.
@Philip,
The specs listed for the DJI Matrix 210 RTK certainly support your statement; although they are listed as hover accuracy, which isn’t really what we need.
Vertical: ±1.64 feet (0.5 m) or ±0.33 feet (0.1 m, Downward Vision System enabled)
Horizontal: ±4.92 feet (1.5 m) or ±0.98 feet (0.3 m, Downward Vision System enabled).
I’d like to see RTK GNSS spec sheet for the 210.
4.92ft is really poor performance for an RTK equipped machine.
2 Likes
@ Antionio,
I don’t know why they photos wouldn’t benefit from precision movement. If the drone flies with better GNSS data, I am sure the photos would be tagged with the same information.
Dave
@David
I wonder why you would need such hover accuracy if you are using the Z30 which is designed to be used from a safe distance from any structure.
It seems that DJI’s main selling point for the RTK system is a way to get around any signal interference from antennas etc.
I agree that the stats are pretty poor…if I was paying that kind of price I would be expecting a fully survey grade machine with centimetric accuracy.
Hate to say it but early adopters be aware…been there myself with another over hyped quad that came out a couple of years ago…
@Philip,
The need for hovering accuracy is dependent on the application. One of the reasons we purchased an RTK GNSS receiver was to improve hovering performance, particularly in the z axis. I’ve never done any rigid testing to validate this claim, but empirically, our quad hovering accuracy is probably less than a foot vertically. We need this for certain applications.
Dave
@Dave
I use a survey grade fixed wing with both RTK and PPK. Base station is a Topcon HiPer SR.
So are you using the DJI M210RTK with a DJI base receiver or another base?
@Philip,
Our RTK platform is based on the pixhawk. We typically get to an RTK fixed solution by sending RCTM from CORs. When we don’t have access to CORs, we use PPK. I don’t know the specific model of our base, but it’s survey grade. My partner is a long time surveyor and was an early adopter of GPS surveying gear. The base is probably a trimble.
@ David,
I agree with you. I think that DJI Matrice 210 RTK is not the solution for an accurate survey but only a way to get around any signal interference from antennas. If so it is too expensive and not risolutive for the purposes of an accurate survey
Take a look at this webinar regarding survey grade RTK/PPK. It’s designed for the Inspire and Phantom 4 Pro, but they say they are testing it on the Matrice. It looks like a good solution.
https://www.expouav.com/session/technical-seminar-high-accuracy-mapping-with-dji-phantom-inspire/
The real reason why we bought the 210 RTK really had nothing to do with positional accuracy. Recently we went through a vetting process with a power company to use UAVs as another survey tool, and they will not allow any overflights of substations as well as mandating 20 foot separation from any of their structures. We intend to do inspections and radiometric work as well as provide mapping and modeling over their entire network. Having the RTK will help mitigate the EMF interference with the internal compass. That was their main concern…the safety aspects of it. So for us, having the collision avoidance, ADS-B In, modular cameras and sensors, weather resistance, RTK, etc. all helped us make the decision to get the 210. It also made our vendor more confident in our ability to operate safely around their infrastructure.
I agree the term RTK is misleading when used in drones (DJI specifically). Unless you are providing some real time (RT) correction via radio link with the drone, I don’t see how you can truly get absolute accuracy. We always use surveyed ground control and supplemental data as a QC. I don’t think there is any other way around it at this point.
Pix4dCapture has been our go-to app for field collection on the mapping side. It really is easy to use and easy to train others how to use it. If we need to look elsewhere for capture software, I guess we will have to do that. I just hope that is not the case.
2 Likes
Dave,
Your reasoning makes perfect sense to me. For similar reasons we are thinking about the 210 for bridge inspections.
After some time in the field, I would love to hear your opinion of the 210
Dave
3 Likes