my question is more about what deliverables actually make sense here. like do i just hand them jpegs or is there a proper workflow for this. i have a mini 4 pro and was thinking about picking up a used air 3 but honestly not sure if the camera difference matters much for roofing versus just getting good coverage angles. some of the houses they do are in pretty tight neighborhoods so there's that too.

also do i need anything special in terms of insurance or contract structure if im doing this for a business rather than just one-off real estate stuff. never done repeating commercial contracts before

been thinking about the matrice 350 with the H20T or even just renting a zenmuse H30 setup for the jobs that need it but the cost is brutal. also not sure how much difference it actually makes in practice vs just flying tighter orbits and accepting the increased risk near the structure. anyone doing this regularly with a workflow that actually works well for the deliverables clients expect?

we do telecom tower inspections, cell towers mostly but some broadcast stuff too, and we've been pitching drone-based inspection to a few clients as a way to reduce the number of actual tower climbs. safety argument is obvious, cost argument is obvious. but keep running into the same wall where the structural engineers reviewing the reports dont fully trust the drone data for things like bolt torque verification or antenna mount wear.

i get it, there's stuff you literally cannot assess without hands on it. but we're using a zenmuse h20t on an m300 with a 23x optical zoom and the image fidelity is genuinely insane, like you can see corrosion on individual bolts from 5 meters away without putting anyone in the air at 200ft. the issue seems to be more about documentation standards and nobody agreeing on what constitutes an acceptable drone inspection report for structural sign-off.

anyone dealt with this? is there like an actual standard emerging for this or are we all just making it up as we go

the camera is obviously incredible and i can see stuff i would have missed from a ladder, like flashing issues around chimneys and soft spots you can kind of tell by the way shingles are sitting. but the actual workflow is what's killing me ? i'm using pix4d for the orthomosaics and it takes forever to process and then i still have to manually go through the photos anyway. so like where's the time savings exactly.

the other thing is liability. had a homeowner ask me if my drone footage would hold up if there was a dispute with the insurance company and i genuinely had no idea what to say. anyone else doing roofing inspections professionally and figured out a good workflow that doesn't require a computer science degree

it works but its not super professional looking and some of the larger property management companies ive been pitching to want something more structured. one of them asked if i could provide measurements and im not sure if the photogrammetry output from my setup is accurate enough to quote linear footage of damaged flashing or whatever with any confidence.

also the turnaround time is kind of killing me - processing takes forever on bigger commercial roofs and by the time i get the deliverable to the client they've sometimes already had someone up there manually. anyone streamlined this? wondering if its worth paying for the higher DD tier or if theres a better processing pipeline people are using. pix4d? metashape? just looking for what people who do this regularly have landed on

the thing is our clients are pushing hard to cut climb costs and the insurance side of it obviously makes drone-first make sense on paper. but there's stuff you just cant catch without someone actually putting hands on the hardware. loose hardware, weatherproofing tape lifting, that kind of thing. the drone gets you 80% of the way there maybe.

anyone else doing telecom tower work with drones? curious what payloads people are using and whether youre doing full facade scans or just targeting specific sections based on initial visual pass. also wondering if anyone has dealt with pushback from clients who want the drone report to replace the full structural inspection report entirely, because thats a headache we keep running into

my gut says its compass interference from all the steel and the antenna arrays but the controller never threw a compass warning. managed to get it back out safely but i lost probably 40% of the footage i needed. anyone dealt with this before around heavy metallic structures. running a autel evo II pro if that matters

im shooting in 4k and zooming in on the footage after the fact but half the time it still looks ambiguous. like i can see something might be wrong but i cant tell my customer definitively. wondering if anyone else doing roof work has figured out a better workflow or if im just expecting too much from this thing. also the oblique angle stuff ? how steep can you actually go before the footage is basically useless

went back and looked at my footage frame by frame and you can kind of see it if you know what youre looking for but its not obvious at all because of the angle the sun was hitting that face of the chimney. the shadow basically ate the detail. im not trying to make excuses but i also want to figure out how to not have this happen again. do people fly these things twice at different times of day or is there some other approach im not thinking of. thermal seems overkill for roofing but maybe im wrong about that

the thing is we already own a mavic 3 enterprise for some site documentation stuff so its not like we're starting from scratch. but for actual close-up tower inspections i keep reading that you really need something with a good zoom or thermal, and the M3E thermal combo is decent but ive seen people say the zoom on the tele lens isnt quite enough for really tight component inspection from a safe standoff distance. anyone actually running a regular inspection workflow on towers? curious what platform youre using and whether the ROI actually penciled out compared to climbers.

ended up having to request a temporary shutdown window on the equipment which is a whole process with the carrier. some towers we can do fine from a distance, others you basically have to get the thing within a few meters to see the mount hardware and connection points properly and thats where it gets sketchy.

currently looking at whether its worth getting a skydio 2+ for the obstacle avoidance in tight spaces or if the RF issue is just going to make that worse since skydio relies so heavily on the vision system rather than gps. anyone dealt with this specifically or have a setup that actually works reliably for tower work

the thing i wasnt expecting is how much better the documentation is. clients actually trust the assessment more when you can pull up photos and video showing exactly where the damage is. had one guy who was arguing with me about whether he needed new flashing around his chimney and i just pulled up the footage on my ipad right there and he was like oh yeah okay that does look bad.

my question is more about workflow i guess ? how are you guys handling the deliverables? im currently just exporting clips and photos and dumping them in a google drive folder but it feels pretty janky. is there software specifically for roofing inspection reports or is everyone just doing something similar to what im doing

i was orbiting at about 180ft maybe 15 feet lateral clearance from the structure and somewhere around the 3rd pass the aircraft just started drifting toward the tower. not a gust, the air was pretty calm, it corrected itself but only after i went to manual and pulled out. on review the telemetry shows a big compass error spike right when it started drifting.

i know towers mess with magnetometers, ive always done my compass cal away from the site before flying. but this was worse than anything ive seen before. talked to another operator who said some of the newer 5G equipment they're installing throws off interference in ways that older antennas didnt. anyone else run into this? starting to wonder if i need to be in full atti mode from the start on these or just tighten up my RTH settings so it doesnt try to do anything autonomous near the structure.

honestly the footage was fine for flat commercial roofs, could clearly see ponding areas and some membrane bubbling near the HVAC units. but when i tried to get close to a steep residential pitch to check for missing shingles i kept getting weird wind buffeting from the roof edge thermals or whatever, had to keep my distance more than i wanted.

my question is ? is there actually a meaningful difference using something like a Mavic 3 Enterprise or the Autel EVO II Pro for this vs just a standard prosumer bird? like is the image processing different or is it mostly just the support contract stuff that makes it enterprise pricing? feels like a lot of money if the sensor is basically the same thing.

ended up launching from the street and it was fine but i felt like i was winging it. for people who do this regularly how do you actually structure a roofing inspection flight, like is there a standard pattern you use or does everyone just freestyle it based on the property. also what do you use for deliverables, just a folder of jpgs or is there software that stitches it into something more presentable for a contractor who is definitely not technical

right now we're looking at the M3E or possibly renting time on an M300 with the H20T but the budget conversation is ongoing. main use case is visual documentation of mounts, cable runs, coax condition, rust spots on the structure itself. not thermal, just RGB for now. anyone running inspections on telecom infrastructure regularly and have opinions on which setups actually hold up when you're 3 feet from a panel antenna that's definitely transmitting

the issue was the guy wires. i knew where the main ones were, had the site map, walked the base before flying. but theres a secondary set of intermediate guys about 2/3 of the way up that werent on the drawing we were given and i came within maybe 3-4 feet of one on an orbit pass. just caught it visually in time and pulled out manually. heart rate definitely went up.

flying tower inspections in manual mode is kind of essential for exactly this reason, the obstacle avoidance on the M2E (enterprise) just gets confused around all the lattice structure anyway and i had it disabled. but yeah, never fully trust the site documentation they give you. always do a slow visual climb on the first pass before doing any orbits or close approaches.

anyone else do a lot of vertical structure work? curious how others handle the guy wire problem specifically because i dont see it talked about much

the workflow right now is basically fly the perimeter, do a grid over the flat sections, dump everything into a folder and review later. we use DJI Fly still which i know is probably not the right software for this but it works. the issue is when clients want an actual report with measurements or damage callouts we're kind of doing that manually in post which takes forever.

anyone who does this professionally ? what are you actually using? is the mavic 3 thermal worth it for roofing specifically or is that overkill. also wondering about whether we need Part 107 for all of this because our guy has it but we're not sure if subcontracting to clients counts as commercial use on our end

problem is every time i got within maybe 15 meters of the larger HVAC clusters i was getting signal dropouts, not a full disconnect but enough that the video feed was getting choppy and i had a couple moments where the aircraft just kind of hesitated before responding to inputs. im flying a mini 3 pro which i know isnt the ideal tool for commercial work but its what i have right now and it usually handles fine.

i assumed it was RF interference from the units but im not totally sure. wind was calm, battery was good, no other obvious reason. ended up flying a wider orbit to stay clear of them which means i have some gaps in the thermal coverage right over the units themselves which is kind of exactly where you want the data. do i need to go back or is there a way to work closer to that kind of equipment

the towers we're doing are mostly lattice steel, some monopoles, heights ranging from like 60m to maybe 140m. the H20T zoom handles visual pretty well and the thermal picks up hotspots on electrical components which is the main deliverable anyway. but now theyre asking about deflection analysis and corrosion mapping and im not sure if that requires point cloud data or if you can get there with dense photogrammetry outputs from something like reality capture.

anyone actually running lidar payloads on tower work specifically, not just general construction scanning. curious what the real world difference looks like in the data

thing is i got about 3/4 through the flight plan and the signal just dropped. not a full disconnect but it went to like 1 bar and the video feed got choppy and it started drifting. i was standing on the ground at the edge of the building and there were maybe 8-10 large HVAC units on the roof between me and the drone. got it back manually but it was sketchy for a second

my gut says the HVAC units were blocking line of sight to the RC and maybe also putting out some interference. has anyone else ran into this? thinking i need to either fly from the roof itself next time or use something with better signal penetration. or just do the flight plan in sections and reposition. the mini 3 is fine for residential roofs but this felt like it was at the edge of what it can handle

like for a guyed tower checking for rust, loose hardware, bent members ? i dont really see what thermal adds over a good zoom. you can see mechanical damage just fine at 7x or whatever. but then i had a job last month where the client wanted to check for heat buildup in the coax connectors at the top and i was kind of wishing i had the radiometric option

anyone doing this regularly? do you bother carrying the thermal payload or is it mostly dead weight for structural steel stuff

the other thing thats been annoying me is the magnetic interference near the antenna arrays is screwing with compass calibration on my M30T more than i expected. not dangerous exactly but it makes the bird a little twitchy in attitude hold and i end up spending more time than i should managing it manually. anyone else running into this and is there a standard procedure im missing or do you just deal with it. also curious if anyone has moved to a more automated mission profile for towers because the repeatability would really help with the documentation side of things.

thing is im running into clients who want actual documentation for insurance purposes and im not sure if what im capturing is detailed enough. im shooting in regular photo mode at maybe 20-30 feet AGL and it looks fine on screen but when you zoom in on the deliverables the resolution starts to fall apart a little. wondering if anyone doing this professionally has figured out the right altitude and overlap settings to get usable deliverables that actually hold up when someone scrutinizes them. or if theres a better sensor option im missing.

wind was moderate, maybe 12-15mph steady with some gusts. flew around the tower in a kind of manual orbit keeping about 15 feet off the structure. got decent footage of the guy wire attachment points and the top section but honestly i was sweating the whole time because the signal was getting a bit sketchy up high and there's all that RF interference near the antennas. drone held position okay but i was not comfortable and i kept wanting to be further away from the structure than i was.

ended up giving the client usable footage but i told them for anything structural they really need someone with proper inspection software and a more capable platform. just felt weird flying a consumer drone in that environment. anyone else done tower work on improvised equipment and lived to tell about it