Home Inspector in Arizona Looking at Drone Options for Roof Inspections


Motopreserve

Drone Enthusiast
I think your concept is great. And the fact that this "commercial" endeavor has been hatched to keep you safe (on the ground) but still do your inspections is the exact type of thing the FAA needs to see. Good use of the technology that doesn't invade privacy nor rain hellfire :)

But for now, I hope they don't see it :)

Not sure if those linked cameras zoom the way you need - but a lot of people use sony Nex5 or 7 on these quads with good results. Somewhere between a simple gopro and a full on DSLR. You can get gimbals for them - and could get better close ups with the stock lens. Used they can be purchased for about $400 (nex5).

Good luck.
 


Motopreserve

Drone Enthusiast
I'm not sure if there is digital adjustment of the zoom for in-flight adjustments, but you can definitely get closer with the lens than you'd be able to with the gopro.

Its seems you might be able to determine a good focal length and then stick with it for most shots, no? Or would you need to zoom after discovering an issue?
 

Accupro

Member
Just don't know if I would need the zoom or not, just want to try and if possible have it! I have actually landed on the roof with my AR Drone in such a way I could get the shot and then took back off again. There is a ground effect with the AR getting close to the roof surface and then moving around on the irregular angles, not sure if this would be the same with one of these higher-end quads. This was another reason why a zoom would be a benefit in addition to the ability to stay away from trees and shrubs over growing the roof line.

I also tend to take a picture of the manufacturers label on a package AC?heater on the roof when possible but I'm thinking that may not be possible with the drone, its more like a close-up.
 

SoCal Blur

Member
Just don't know if I would need the zoom or not, just want to try and if possible have it! I have actually landed on the roof with my AR Drone in such a way I could get the shot and then took back off again. There is a ground effect with the AR getting close to the roof surface and then moving around on the irregular angles, not sure if this would be the same with one of these higher-end quads. This was another reason why a zoom would be a benefit in addition to the ability to stay away from trees and shrubs over growing the roof line.

I also tend to take a picture of the manufacturers label on a package AC?heater on the roof when possible but I'm thinking that may not be possible with the drone, its more like a close-up.

Have you seen this? This guy is struggling in strong winds with his quad to inspect his roof, but he is using the GoPro and it should give you a pretty good idea what you can expect from using a GoPro: https://www.youtube.com/watch?v=yOErbfutyBM
 

Motopreserve

Drone Enthusiast
Check out THIS product. I think it may handle zoom. I haven't researched it enough.

Keep in mind any camera with any type of zoom at all will allow you to film/photo "closer" while still keeping a safe distance. This may lead to higher quality shots that serve your purposes. So options like a sony nex may give you what you need without having to zoom "on the fly." Maybe not.

There red is also the route of standard def FPV style cameras that can be rigged with head tracking so you could pan the roof shots from the ground. It's a fascinating technology to me. People couple these with DVRs on the ground (or sometimes on the camera) for video monitoring later.

Man, the options just keep mounting :)
 


Accupro

Member
Very interesting and it appears he was using goggles. To me its obvious you would need a gimbal and a tilt option at a minimum and hopefully a remote controlled zoom. I could see were if there was too much wind then it would not be an option maybe this speaks to a hex versus a quad in that the hex would be heavier and thus able to be more effective in a higher wind?

I have done much better than that with my little AR.
 

SoCal Blur

Member
Very interesting and it appears he was using goggles. To me its obvious you would need a gimbal and a tilt option at a minimum and hopefully a remote controlled zoom. I could see were if there was too much wind then it would not be an option maybe this speaks to a hex versus a quad in that the hex would be heavier and thus able to be more effective in a higher wind?

I have done much better than that with my little AR.

Yes, as a general rule, the Hex and Octo versions would be more stable in wind but a lot has to do with the FC. I assume that you're flying your AR without the foam bumper - that can act like a sail in wind and really ruin your day.
 

Accupro

Member
Dude I was probably looking at the same web page when you were, what a coincidence. I also found one in Australia for $90 that does the same thing but the VP-Systems looks a little cheaper and has lots of interfaces and supports most RC transmitters.

It basically offers 4 different interfaces, two different USB stds, IR and LANC, the USB/CHDK method seems the way to go as it supports some lower end digital cameras say in the $300 range (10 - 12 meg). I could see where you could position an IR transmitter in front of the camera but thats just more complication. Looks like the thing to do would be to research which digital cameras support the USB/CHDK standard and find a gimbal that supports the camera.
 

Accupro

Member
Yes relative to the AR. Based on the input on the AR Forum I am using the outside hull and set the software for the inside hull which tends to slow down reactions to the FC and kinda gives you more stable flight. I also pulled all the extraneous plastic logo stuff off the outside hull, cut off a portion of its nose to support a camera swivel, balance the props and upgraded to bearings from bushings. I get really stable flight and good control.

Problem is the WiFi control which limits distance significantly, even with a WiFi extender I get pixelation of the video feed every now and then. Its also drifted out of Wifi range a couple of times but I was able to go get it back, but to me this would not be acceptable for my inspections, would need something more reliable.

With the inspections time is money so while my original idea in using the drone was to allow me to work longer and minimize the risk of falling I was also hoping to maybe do the roof portion of the inspection faster but after trying this for awhile I'm thinking its going to take the same amount of time, maybe even longer. On a standard say 2K s.f house I would probably take 15 to 20 mins. to get out my ladder, set it up, and walk the roof. I'll probably get faster over time with the drone but will probably still take the 15 to 20 mins. with set up and flight time. Were the bigger time saver would be on two or three story condo's and larger commercial properties.
 


kloner

Aerial DP
a competent well seasoned pilot could scan 20,000 feet of roof in a couple minutes 2-3 feet over it. you might be better served finding somebody in your area to give you a smokin deal to scan roofs. if it takes 15-20 minutes to fly a roof it'd be a waste of time and risk..... there has to be reward or using a uav is a waste
 

SoCal Blur

Member
wind is wind no matter what the airframe......

I have to disagree with your statement. Speaking as a real-life helicopter pilot, I know a little about wind and it's effect on airframes. Also, I have a small quad that has a removable foam bumper (like the AR Drone) and in windy conditions, the quad with the bumper will get overwhelmed by the wind and I essentially have little to no control over the quad whereas if I remove the bumper I at least have a fighting chance.
 

Accupro

Member
Good to know, I'm probably assuming it would take me longer with setup and everything then it actually would!! At one point I was thinking I could just quickly scan the roof going back and forth, front to back and then the time consuming portion would be editing the video. I also would not need to worry about having a real clear image of what the camera was seeing, in other words no need for goggles. After I tried that several times I could see for instance seams in a rolled roofed portion that when going in one direction looked like a typical seam but in the opposite direction I could see the seam was open (a defect), differences being shadows and the angle of the sun.
 

kloner

Aerial DP
so-cal, that wasn't directly at your rail comment, it was at airframes and motor counts in general

I engineer uav and test more gear in a month than most people see in a lifetime.... I drive 2 hours each way just to get in the wind 3-4 days a week, i'll disagree with your disagreement based on my observations

there is virtualy no difference from the number of motors. it's strictly the size and the amount of thrust compared to it's weight..... start putting too much weight they all suck, that is a 550 with anything bigger than a hero zen, she becomes overloaded and is a pig in the wind... quad is easy to tote around, hex becomes a baby seat in the car, a octo becomes the size of a couple people... i can fly my 5 lbs disco in wind with better results than i can my 32 lbs octos,,,,, bar none, day in and day out.....
 
Last edited by a moderator:

kloner

Aerial DP
when we show up to a set with a octo, 3 axis gimbal, fpv gear, it takes us 30-40 minutes to setup..... with a one man disco it takes me about 5 minutes to fire off. but seriously, go up high enough to see the whole roof then come down on it and scan along in 1080 60p then just slow the footage down to 24p and file your report, collect the check. But to start out thinking your gonna go from an ar drone to flying any real mr over roofs incident free would be a pipe dream. Your gonna hit wires from antennas, chimneys, the neighborhood tom cat.... stuffs gonna happen. it does to all of us, the idea is to make it as safely done as possible, if that's too challenging sounding, let somebody else do it for you.... seriously, some kid with some stick time on a phantom is probably ready, craigslist, reader, etc. have him come over and scan your roof..... see how long it takes and figure a value you can agree on. Give yourself time to get that good practicing on yours then jump in. doesn't mean you can't own one at the same time, but to reliably do work with them needs precission cause your not gonna get parts in a day and have it right back up. your gonna have a rig down and your back on the ladder with a bad taste in your mouth now,,, frikin multi rotor

Good to know, I'm probably assuming it would take me longer with setup and everything then it actually would!! At one point I was thinking I could just quickly scan the roof going back and forth, front to back and then the time consuming portion would be editing the video. I also would not need to worry about having a real clear image of what the camera was seeing, in other words no need for goggles. After I tried that several times I could see for instance seams in a rolled roofed portion that when going in one direction looked like a typical seam but in the opposite direction I could see the seam was open (a defect), differences being shadows and the angle of the sun.
 

Accupro

Member
So just doing a quick look around my current Canon ELPH 300HS 12MP which I currently use for inspections supports USB/CHDK and this combined with the VP-Systems RC interface might do the job and the weight is 142 grams. I could potentially "see" what the camera sees in a pair of goggles (say Fat Sharks) and allow the SD card in the camera to record or take pictures of what I want - coooool!!

For me I use the lowest resolution available on the camera as I am incorporating relatively small pictures in a PDF inspection report so high resolution pictures are not important but high resolution goggles would allow me to pick out the detail.

Kloner do you have any experience integrating the new Vuzix HD goggles (http://www.vuzix.com/consumer/products_wrap_1200dx/) into a UAV system?
 

tstrike

pendejo grande
when we show up to a set with a octo, 3 axis gimbal, fpv gear, it takes us 30-40 minutes to setup..... with a one man disco it takes me about 5 minutes to fire off. but seriously, go up high enough to see the whole roof then come down on it and scan along in 1080 60p then just slow the footage down to 24p and file your report, collect the check. But to start out thinking your gonna go from an ar drone to flying any real mr over roofs incident free would be a pipe dream. Your gonna hit wires from antennas, chimneys, the neighborhood tom cat.... stuffs gonna happen. it does to all of us, the idea is to make it as safely done as possible, if that's too challenging sounding, let somebody else do it for you.... seriously, some kid with some stick time on a phantom is probably ready, craigslist, reader, etc. have him come over and scan your roof..... see how long it takes and figure a value you can agree on. Give yourself time to get that good practicing on yours then jump in. doesn't mean you can't own one at the same time, but to reliably do work with them needs precission cause your not gonna get parts in a day and have it right back up. your gonna have a rig down and your back on the ladder with a bad taste in your mouth now,,, frikin multi rotor

this is like gospel, well said dude!

Accupro, you'll need a live out on any camera you choose to use. Something like the rx100 on a simple servo based gimbal would probably do the job. I've even seen servos used to control the little zoom button and shutter but that gizmo moto referred to is the way to go. Even at 20-25 minutes, I'd rather lift a quad into my van than an extension ladder on the roof rack. I've got my quad set up down to under 10 minutes from case to air with live video out of the gopro down to a 8" monitor. No goggles for me, like to see on my periph's.
 

Top