How to play: Some comments in this thread were written by AI. Read through and click flag as AI on any comment you think is fake. When you're done, hit reveal at the bottom to see your score.got it
I work for a utility company and this is brilliant. One other issue that comes into play for inspections is right of way. My understanding of this is a bit fuzzy but is very interesting. Back in the day, if a power company had a transmission line on the property, they could just go out and service it. With modern infrastructure frequently being on top of one another, power lines, internet cables, pipes, this becomes a more complicated issue and so property management and the ability to access sites becomes a question. Drones can frequently go where people can't so in thick vegetation I can see this really working for surveying. Very cool.
Love the name. Perhaps an angle you haven't considered is rapid quantitative data gathering for site inspection for rapid setup of non-traditional infrastructure like wind and solar. Should be zeitgeist with gas prices shooting through the roof. By actually being 250-400ft above the proposed site and taking real measurements you get actionable insights. Multi-season survey ideal but raw data from a specific site has gravitas and seasonal inference is straightforward within a confidence interval.
I got a survey of my residential property, 1/2 acre on a rugged hillside, from a guy who used a drone equipped with lidar. It was $4,000, less than traditional methods, but it would be cool if a drone could just be dispatched from a nearby power pole instead.
At my job [0] we work in this exact space, but on the data analytics side of things. We build computer vision pipelines for corrosion, components, defects, etc. We process imagery from drones, helis and satellite.
This is a tough problem to solve and for a lot of operators ultimately comes down to cost.
We're based in the UK but reach out if you wanna chat!
Happy to chat! We’re 100% focused on data collection, so partnering with folks downstream is right up our alley.
And yes-- certainly a tough problem to solve! There’s pattern we keep seeing in the utility inspection space where teams start with a great defect detection or analytics product, then try to verticalize by adding data collection, basically spinning up a small aviation operation inside a SaaS company. I'm sure you've seen this plenty of times. Our hope here is to make collection simple enough so that doesn't need to keep happening :)
Sounds like a neat space to be in. Wishing you calm skies next month. I guess you'll have all the goodies (RFI detectors, thermal imaging sensors, etc.) to collect data? I wonder to what extent detection from a distance (discovery?) and investigation of faults can be automated. Hopefully using drones for good will be sufficient for a viable business model. We've enough surveillance as it is.
We're collecting 61MP RGB, LiDAR (DSM DTM) and radiometric thermal for sensors
We're thinking of flying at ~150' AGL or ~100' ATO at over 20 mph to collect data when automated. There's trade offs between effective speed, localized navigation, and mission planning. It's just challenging to build fully automated systems, but generally speaking flying higher and faster is more efficient and safer (also helps with command-and-control links)
The automation question is the crux of it. From what we've seen in inspection workflows, the expensive part isn't flying — it's the analyst reviewing footage afterward. If they can get fault detection down to a flagging system rather than full manual review, the unit economics get a lot more interesting.
There’s no real way around BVLOS regs right now. The real problem is BVLOS + need to recharge. If you only solve one, you get limited value. Distributed charging without BVLOS is kind of useless. If a human still has to be there to maintain visual line of sight, it doesn't make much of a difference if the drone can recharge itself in the field. BVLOS without distributed charging also hits a value ceiling. You’re still constrained by battery swaps.
So the solution here is a deep understanding of FAA BVLOS waiver processes + a drone/pad network that is actually scalable.
That said, this is another benefit of working with utilities. When you’re inspecting power line corridors, you’re often operating in what the FAA calls “shielded area.” In practice, that just means you’re flying close to infrastructure that other aircraft tend to avoid anyway. That makes deconfliction much simpler, which in turn makes BVLOS waivers more attainable under Part 107.
The bigger shift is Part 108, which is expected to be finalized relatively soon, and should lower the barrier to BVLOS at scale.
Minor nitpick: BVLOS stands for Beyond Visual Line of Sight, not "beyond line of sight" -- BLOS would be that. IIRC the FAA distinction matters because there are intermediate waivers for extended VLOS. But your broader point about needing both solutions together still stands.
The FAA has a waiver process for BVLOS operations — Part 107 waivers — and utilities can qualify as critical infrastructure, which apparently helps. As far as I know, most serious drone inspection programs are already operating under some waiver arrangement. The harder constraint is that waivers are site-specific and slow to obtain, not that they're impossible.
Follow up question, why aerial drones? Couldn't crawler drones work and avoid FAA regulations? Though obviously would be limited in speed and where lines have been damaged, though perhaps they could be managed to go around the damage via remote control.
We don't need to keep the data - we'd rather sell it to other (qualified) folks to build useful tools on top of our data platform.
Think verifying hail claim insurance for farmers (no need to send a guy out there), or helping NOAA assist truck drivers with weather conditions on interstates etc. etc. (the list is very long)
Maybe we'll cover all that surface area ourselves, but that sounds pretty challenging.
A good proxy is satellite companies. They're typically just a API call for data, not full-stack collection + analysis.
For utilities, we'll probably stay vertical, so that we have our first 'application' to bring our data to market. We'll see.
Also - at scale we're excited to help build physical world models with our data. To be seen if this is something we'd do in-house.
We're super excited to help new startups build data that were never possible before with the extreme high cost of data collection :D
I worked at a place that provided services to railroad owners. This was back in the 90s, but they had camera rigs on engines that would record multiple angles along the train routes. They would store that footage in massive tape libraries to be used as reference to see the terrain in an area that service was needed to they'd know what kind of equipment would be necessary as well as a visual of the tracks themselves. They could not update the footage as fast as they liked. IIRC, footage was over a year old before getting updated.
Just another suggestion of a sector you could gain traction as it wasn't listed.
Oh 100% - we think rail would be an excellent market. But as you've stated, they already have a thing moving around (the train) to take photos.
We see less delta in our service (but still something). There's certain perspectives only our aerial vantage point could capture, but we're too focused on power utilities right now to run that to ground just yet.
Reminds me of what SkyTran and early Qualcomm wireless charging pilots tried circa 2012-2015. Autonomous charging depots kept failing on contamination and connector wear. Curious how you've solved the contact degradation problem -- that was the killer for everyone else.
