Welcome, Commercial Drone Pilots!
Join our growing community today!
Sign up

Cell Tower 3D Models

I suppose the challenges may include the following:

Having the proper equipment and keeping your equipment flying safely around the structure, expecting GPS loss and lost links.
Knowing what the client is trying to see and what the data will be used for.
Gaining access to the tower with permission to fly very close to it.
Avoiding hitting guy wires and flying over busy streets that may be very close.
Airspace limitations.
Appropriate insurance.
Knowing who owns the tower and or carrier assets that are on the towers.

I could think of a few more and knowing your intentions could help with better info to you.
 
I suppose the challenges may include the following:

Having the proper equipment and keeping your equipment flying safely around the structure, expecting GPS loss and lost links.
Knowing what the client is trying to see and what the data will be used for.
Gaining access to the tower with permission to fly very close to it.
Avoiding hitting guy wires and flying over busy streets that may be very close.
Airspace limitations.
Appropriate insurance.
Knowing who owns the tower and or carrier assets that are on the towers.

I could think of a few more and knowing your intentions could help with better info to you.

Thanks for listing these! Definitely very insightful.

With regards to the data capture though, with the aim of generating a 3D model later, are there specific parameters that you configure the flight with?

For instance, so far we've learnt it's useful to:

- keep the sky out of the pictures
- use an oblique gimbal angle (facing downwards) while going around the tower
- use high overlap (~85%) between the images

We're wondering if you've observed additional configurations that help?

Our goal is to build flight planning and automation tools that help in capturing high quality data for a number of scenarios - cell towers being one of them :)
 
Most clients will need at least 1 angle pointing upward into the sky for infrastructure imaging. Lots of components can only be seen using upward angles. Some of those angles will need exposure input from the photographer/pilot to get a proper image. Auto/program exposure mode usually fails in those scenarios. The sun is not always in the optimal position, I prefer when it is overcast for consistent imaging.

Most of the time the client will dictate overlap which is usually protected by an NDA :)
 
Most clients will need at least 1 angle pointing upward into the sky for infrastructure imaging. Lots of components can only be seen using upward angles. Some of those angles will need exposure input from the photographer/pilot to get a proper image. Auto/program exposure mode usually fails in those scenarios. The sun is not always in the optimal position, I prefer when it is overcast for consistent imaging.

Most of the time the client will dictate overlap which is usually protected by an NDA :)

Thanks for sharing more on this! We completely agree with regards to the exposure settings, it's critical that the optimal parameters are set before every flight to ensure consistency across the image dataset.

We're curious to learn why the image overlap is protected by an NDA?
 
  • Like
Reactions: SanCap
Thanks for sharing more on this! We completely agree with regards to the exposure settings, it's critical that the optimal parameters are set before every flight to ensure consistency across the image dataset.

We're curious to learn why the image overlap is protected by an NDA?

Short answer is that I worked for a company that paid me to fly cell towers on their contract. They gave me a prescribed shot sheet and flight profile. I consider that their info that they figured out how ever they did and I will not share that info.
 
Not problem at all, and that makes sense. We're currently exploring conducting our own set of trials on a cell tower and hope to publish a short guide on the best practices for 3D models. Thought we should also reach out the larger community to learn more!
 
  • Like
Reactions: SanCap
Not problem at all, and that makes sense. We're currently exploring conducting our own set of trials on a cell tower and hope to publish a short guide on the best practices for 3D models. Thought we should also reach out the larger community to learn more!

I look forward to seeing your results.
 

Members online

Forum statistics

Threads
4,288
Messages
37,643
Members
5,984
Latest member
jaklein91