Replies to This Discussion

Permalink Reply by Arpad on Thursday

Hi, Very very interesting challenge! This job is giving a lot of additional value to why people should develop things like drones, or equipment that drones can carry.

To better understand your project, I would suggest to possibly provide more details on the job. I understand there are several storage buildings, which have same or individual sizes, a flat floor and flat ceiling and flat or cylindric walls (or any other defined shape), those are filled partly with a grain pile that has a non-flat surface shape on the top.

The buildings are what size, what shape? I understand those are not like towers.

The surface of the grain pile is a complex shape? Or, can it be assumed a simple math 3D curve like parabolic or something? (In my mind, when all grain is poured in from a single point, it becoming the top of pile, and the surface is relatively a simple 3D curve)

Does the pile always meet the side walls, so there is a certain height at the walls? If you just need to measure the top height and the grain heights at the side walls, then it is probably easier task (vs accurate 3D mapping). Just need to do some math for the volume calculations, based on these known parametres.

Alternatively, are you aiming to 3D-map the surface? Seeking some photograpic or ultrasonic or other means of measuring / 3D mapping? Do you want to achieve certain accuracy?

Let me jump to some fundamentals.
What is your intended achievement by using the drone? Do you just want to cut walking time, just replacing human walk with drone faster accessing destinations? Sure, a drone can view from above without climbing up stairs.

Or, is it your aim to increase accuracy a lot? Needless to say, 3D mapping is beyond scope of human walk.

In addition, I believe there are a few things you are considering besides focussing on measuring.

Room of manouvering for drone seems pretty small (1m). Technology should assist height/distance stabilization of the copter.

When flying so close to the grain, dust might be an issue for your sensitive aircraft, lifetime of the motors, or clarity of the camera lens can be affected.

Potentially obstacles down from the ceiling or from walls, like lamp/vent ducts etc. might be another issue.

Just thinking "loudly". Please tell me if I'm raising too many questions, as you mentioned you are already one month in the project :-)

• ▶ Reply

Permalink Reply by Arpad on Thursday

Hi, it's me again. Just seen this new blog (below link) that might be of inspiration. Overkill though, far far more than what we need here. Just imagine to put that cam on your drone. Theoretically... would do the job, eh?
Okay, I know, you need a workable solution...

http://diydrones.com/m/blogpost?id=705844%3ABlogPost%3A2390383

• ▶ Reply

Permalink Reply by Logan Torrance on Friday

To add some clarity, the buildings are roughly 600ft long, 250ft wide, and 35ft high at the center with sloping walls down the side.

The goal: We want to accurately and efficiently measure our grain in these buildings.

Right now: We have crews of men walk on top of these piles and measure the distance from the top of the pile to the roof. From there we plug in the numbers and we have a system that calculates the bushels.

Why a drone: The idea of the drone came because accuracy has been a slight problem, and we have a Phantom 3 Professional. We are flexible with the equipment we use. We looked into 3D stuff, but its a tight space. 

I hope this helps, if anyone has any other ideas, or questions please don't hesitate to ask. As of right now, the FDA and all other flat storage companies have people walking buildings. This could be very big for the ag industry if we can get something figured out!

Also, the pictures were taken by our phantom 3, just gives anyone some clarity to what we are working with.  

• ▶ Reply

Permalink Reply by Arpad yesterday

Hi again. Just in case you do not find a ready made tool. You could put two ultrasonic metering sensors on the drone, one facing down, the other one facing up. An onboard microcontroller (eg. Raspberry Pi) could read both sensors and log the data on an SD card, together with current gps position of the aircraft. Take enough points while piloting drone from point to point in desired density of a net above the surface of the entire pile. A software on ground unit shall put data together, build a 3D map, and calculate the pile volume. The result could be pretty accurate, even if you avoid the peak of the pile.
Optionally logging the Gps height may help subsequent data validation as you should eliminate those upper reads affected by beams under the roof.
Ensure testers are vertically aligned by halting aircraft at time of measuring or by using a hinge.
What do you think.

• ▶ Reply

▶ Reply to Discussion