I’m the co-founder HUVIAiR Technologies, a UAV services start up based in Bengaluru, India. We just got confirmation of our first project today. The requirement is to map a 850 acre coffee estate. The customer has the following specific requirements -
To clearly mark the boundaries of his estate on the orthomosaic, which will help him get a precise calculation of the area of his estate.
To calculate the area that is currently being used productively (i.e. area occupied by the coffee plants).
Count the number of shade providing trees interspersed between the coffee plants.
I have a DJI phantom 3 pro with three batteries.
I have a few questions -
For point 1 above, what’s the best way to mark the boundaries (he has fences on the ground but they are unlikely to be easily seen on the orthomosaic). May be I can attach some kind of bright markers along the fences? Do suggest.
No questions for point 2 above. Any suggestions on how we can automate the calculations in point 3 (counting a specific type/types of tree species)?
How do we plan the flights so as to cover the entire estate without missing out any part(s) of it? Also, what’s the maximum area that can be covered in a single flight with a fully charged battery?
The estate is in a remote area with very poor internet. Is there a way I can plan the flights off site and then completely run the mission offline at site? When I click on “start ortho map” or “start agriculture” on the dashboard, I don’t get locations from India.
The estate is on a hill slope. What’s the ideal point for the drone to take off and at what height should I fly it. Is there a formula I can use? Are there any other settings I should adjust on the app, to ensure a good capture of the estate?
I don’t have a near infra red cam for NDVI, but is there any information I can provide him about the health of his plantation with the NDVI produced from the default RGB camera on the phantom 3 pro?
Is there any other analysis I can do for him that can prove useful?
Hello Vikshut, thanks for your time on the call earlier—here are some answers to your questions:
Bright markers may be useful for this. You may be able to see that they are there on the orthomosaic or even the elevation model. You can try adjusting the Elevation model with the Toolbox to make it stand out more.
We are hoping to integrate automatic counting algorithms sometime in later 2016 so stay tuned! Unfortunately, I don’t have any recommendations of other providers that could do this for you at the moment.
We’ve seen people cover 80-120 acres in one Phantom 3 Pro battery at around 350-400 feet AGL but wind conditions and meteorological conditions will affect that. We do have a feature in our mobile apps that allow for “Continue missions”—basically using multiple batteries to continue one large mission. More info here: http://support.dronedeploy.com/docs/flying-a-larger-area-with-multiple-batteries
Yes, you can use the Sync feature for planning on desktop/web browser and syncing to your mobile device. You’ll have to find India manually for now but we are working on a fix for international locations. @chasemgray I think is working on this.
I would likely take off from either the highest or mid-point. You only need to take into account the height the drone is flying at (adjustable in our app) and determine that it won’t crash into any obstacles or the hill itself! The height above ground level (AGL) will be relative to the point you take off from. So if you take off from sea level (0 feet AGL) and fly at a height of 175 feet AGL but there is a huge hill in your flight path that is 200 feet AGL, you will likely hit that hill. Think through this ahead of time to plan accordingly.
You can tell a lot from just RGB imagery. You’d be surprised what a farmer/grower could tell from having high resolution aerial imagery nearly “same day”. It can be quite valuable to them but you’ll only learn from experience.
We are working on getting some analytics that could be applied to RGB imagery into DroneDeploy but won’t have more than what we currently allow for another few months I’d assume. Right now you’d have to do some trial and error to see what your client finds valuable about the imagery.