Phantom 4 Pro drone flight setup with Metta Technologies' PPK KIT PHANTOM4 and Tuffwing's SNAP PPK Kits

[juliodurand (IDR Industrial Drones & Robots)]

Hello, for people who have experience with the Reach M+ and M2 modules I present these operating tips when using the M+ and M2 with the DJI Phantom 4 Pro for PPK capture and processing, using the PPK Kits from Metta Technologies and Tuffwing .

Taking into account that today the maximum frequency of capture or recording of positioning of the M2 module is 10Hz (10 records per second).

Based on this feature, to achieve maximum precision in capturing positioning data during drone flights, we can recommend the following:

[1] Select the highest possible frequency according to the number of GNSS systems you require. For example: In the M2 module: GPS + GLONASS, frequency 10Hz.

[2] So far this is the best flight mode option: “Hover & Capture at a Point” (Speed ​​can vary between 3m / s to 15m / s).

[3] If you are flying in the “Capture at Equal Distance Interval / Time Interval” modes, the speed of the drone should not exceed 5m / s. Remember that at higher flight speed the precision range may decrease, therefore the speed should be as low as possible according to the availability of operating time and the number of batteries.

[4] Regarding the flight schedule, it is recommended to fly in the hours when the Sun’s height above the horizon is ≥45 sexagesimal degrees.

[5] Image overlap: According to the type of terrain.
Flat terrain: Longitudinal: 60%. Side: 20%.
Undulating terrain: Longitudinal: 70%. Side: 30%.
Mountainous terrain: Longitudinal: 80%. Side: 50%.

[6] In any case, at least 3 control points must be placed on the ground in order to readjust the final processed data.

Please, I ask for your opinion based on your experience with PPK surveys and processing. Thanks.

Links:
https://mettatec.com/en/online-store-unmanned-vehicles-drones/
http://www.tuffwing.com/products/Phantom_Reach_PPK_Integration_kit.html

[michaelL (Michael Lambert)]

1. From our testing there is no difference in accuracy between 5 and 10hz so there’s no reason to double the size of the logs. Local bases and RTK networks usually only run at one Hertz so there’s no reason to try to get corrections much faster than that.
2. This is really dependent upon the type of camera. While stopping does increase the sharpness of the image we again have seen no significant increase in accuracy and have never needed to stop at shots.
3. Another one that is dependent upon the camera. We run The Yuneec H520E RTK and have no issues with a 1.5 to 2 second capture time at 16 mph. We do not go above that.
4. if you are capable of photo editing you can fly at 30° with no consequences.
5. Anything below 50% does not provide enough images per point to recognize adequate tie-points for reconstruction. Our default is 75/65 and we do not deviate from that more than 5%. Are you referencing nadir or oblique images?
6. This depends on the size of the site but in general I agree with three points. Four at the perimeter of the site is better because it locks down the triangulation to a flat plane but this is more for horizontal purposes and absolute accuracy.