|
|
Thread OP
|
Yup it seems we had the exact same idea as them and finished our prototype 6 months ago.
We ditched the concept because carrying targets around sucks... And going back to get them sucks too ![]() You can see our prototype triangle at the beginning of this thread. You also can't do as many points as you want and can't do check shots with it and how do you set up over a known point? Again we are surveyors... we know that your clients will ask you tie into control... we give you that ability. It appears we have been paralleling development as I just saw this a few weeks ago. We are also Canadian based so cheaper shipping and support. Steve |
Last edited by saabguyspg; Sep 08, 2016 at 09:28 PM.
|
|
|
|
||
|
Quote:
|
|
|
||
|
||
|
Quote:
Where do you live? I live in the Netherlands so I'm bound to use either NLR (Dutch) or EuroUSC (UK, but recognized by the Dutch FAA). We chose EuroUSC. Without the certification, we can not legally fly commercially, since the M600 is over 4kg and because we also want to enjoy the options possible when having a ROC (RPAS Operating Certificate). |
|
|
||
|
|||
|
Quote:
Quote:
|
||
|
|||
|
|
Thread OP
|
OK I will do my best but can't give away all our secrets
![]() Is there anything out there in common use that has an update rate fast enough to both be taking photos while moving AND get 1cm position accuracy? Even at 10Hz, moving at a leisurely 10M/s will mean positional offsets as much as a meter. I don't know what people do to address this. I can imagine stopping, taking a photo, moving to a new position, etc., to get the best position information (even then, what position is recorded -- puck location, camera location, etc. -- has to be specified). We do many things to counteract these effects. The most obvious is a true survey grade GPS receiver onboard.. not Knockoffs... running a high refresh rate. This is NOT cheap.... seconds is calculations which I am not allowed to talk much about. We fly about 12+ meters per second. I haven't been keeping up with the DJI RTK discussion, so maybe this is all answered here. But, at least from DJI's website, I still don't know what the system can do. nope they won't tell you anything. That's why for now we won't sell it with our rigs. How can we support our users if all we do is refer questions to DJI and they don't answer them. This DJI is famous for and it will never change. Just for example, does the ground system have its own network RTK capability (i.e., like a survey grade GPS system) to provide it with absolute position accuracy, or do you need a separate GPS device to locate the ground station (and enter coordinates somehow?). Good question. You could definately use their base to set up over a known point had they designed the base to do so. But as it stands you can't. If you are familiar with surveying you know we set up with expensive gear to get mm accuracy at the base over a known point and tie to control... it's not possible with that plate they have so you will not get repeatable data. Repeatable data is worth $$$ On the air side, is inertial data being incorporated to improve recorded position in EXIF, and if so, what's the effective accuracy? If not, does the system record the most recent GPS fix, does it forecast current position based on the most recent vector of flight, etc.? And what is being recorded? Is it the puck location, the CG location, the camera sensor location, etc.? As for DJI I don't know about their GPS but we know the phase center of our antenna EXACTLY and calculate to the camera exactly then calculate offsets in our software. Finally, as a practical matter, does it make a material difference for aerial mapping using photo imagery? What is gained over, say, 3M accuracy and a set of well placed 1cm accurate ground control points? I can see wanting 1cm aerial accuracy if one were parking the rig and doing a 360 LiDAR scan. But with just photos, do you really gain anything? YES you gain huge. No matter what anyone says maps and models made without ground control or geo-referenced photos are inaccurate locally by anywhere from 10 inches to 10 feet or more... it depends on the Z differences one site, overlap and the GPS quality at the time of flight. Take a look at Maps made easy, drone deploy, Agisoft Photoscan etc and all of them will tell you that the maps are inaccurate without GCP or perfect Geolocations If you use ground control you will have the most accurate data close to the GCP. As you get further away data degrades. Anything over 150m data will become pretty useless. As and example we laid 27 GCP's for a survey today. The site was about 300m x 700m Why lay so much GCP? because you have to adjust GCP to the required output. We wanted as accurate as possible along a roadway so we laid out more GCP. Between all those GCP we laid check shots and they have to be defined as such and not GCP's in the software otherwise you are "cheating" and the software will adjust the model to suit the check shot. NOW if you are flying direct geo you only survey say 5 check shots, fly and your data is consistently good across the site. Typically a direct geo rig can take a 6 hour job and turn it into a half hour job with better accuracy. However now you have to trust your system.... All of this is apart from the relative ability to hold position using an RTK system. And, if I were guessing, I would say this is what the DJI system likely delivers -- much better ability to stick a position when the sticks are centered. That may be all it does, but that's still a nice benefit.[/QUOTE] True it is a nice benefit for some.... but for mapping... not useful at all. |
|
|
Thread OP
|
These are the first photos ever released of stryder publicly. This is close to the final production version with minor tweaks like moving the lightbridge out onto the arms and final production "Feet"
We hope you will see much more from this surveying / inspection machine in the future. Steve |
|
|
|
You can set up our base station over a know control point and input that data into the unit. Trust me I'm not expert in surveying and all I have seen DJI advertise is precision position and no compass interference. So I think people might be reading into the capabilities a little more but I'm not 100% sure though since all my job was is to make sure nothing wonky is going on with the FC and that I was seeing the flight precision that I would expect in flight.
I do know the D-RTK does hold better than anything I have seen... We even had a mapping company say they have never seen a copter hold altitude during the entire mission as good as the 600 w/ RTK, +\ - .2 meter, even at my high altitudes of 6900'. |
|
|
|
||
Thread OP
|
Thanks for coming over and joining the discussion Blade we welcome your posts here.
I am interested to see how you can set the base up over a known point. Essentially to work with surveying monuments you will need to bolt it to a tribrach or drop it exactly over a known control point (OBS style) I think it can be done on a tribrach with a special plate but then you will need to input an offset between the DJI GPS antennae and the tribrach as well as a height of instrument over the control point which would be bottom of the base plate to the top of monument. I think OBS will be useful for people to add check points to any survey data created with M600 so I think there is something we can do to help your clients when all is sorted out. I agree, DJI never claimed to be entering into the accurate mapping using direct geo. They have always claimed the accurate flight and that is quite a cool thing in itself. Quote:
|
|
|
||
|
Quote:
You would have to input values in the assistant while connected to the bs, add or subtract the z value of the center point of the antenna. You can also put the base station over this point and press the update button and see how close this is to that known point. |
|
|
||
|
|
Thread OP
|
The ability to do this in your second point is a good thing, then you can adjust if needed.
Keep us posted on progress. |
|
|
Thread OP
|
Hi everyone I have updated the first page.
Most notable OBS will be dual frequency, this reduces wait times at GCP's to 3 minutes or less, yes this puts the price up. We aim for <5000 USD OBS will also act as a base station for Stryder Geo thus you can get OBS now and already have a base for when you upgrade to Stryder Geo. This will be the most complete and accurate drone surveying system on the market. Steve |
|
|
Thread OP
|
Stryder in action today, weather was very windy.
I put up the phantom to get some video but it was terrible, blown all over like a mosquito and video link was horrible. Up went Stryder and it was easily the smoothest and best video I have ever shot. That extra weight and a3 controller are rock solid. Also LB2 had zero issues and much less lag. Steve |