Skydio Live
+00:00 GMT
Sign in or Join the community to continue

Goodbye Fear: Successfully Navigating Drone Program Regulations

Posted Oct 01, 2024 | Views 325
# Drone
# Government
# Regulatory
# Remote Operations
# Utilities
Share
speaker
avatar
Daniel Jenkins
Regulatory Program Manager @ Skydio

Regulatory Program Manager working on BVLOS permissions with Skydio.

+ Read More
SUMMARY

Unlock the future of drone operations with Skydio's groundbreaking approach to Beyond Visual Line of Sight (BVLOS) operations. As the utility sector rapidly adopts drone technology, mastering aviation regulations is crucial to fully realizing the benefits of autonomous operations. Expect to hear about:

• The Journey to Autonomy: Explore the shift from BVLOS operations requiring Visual Observers to fully autonomous drone missions. • Regulatory Milestones: Discover Skydio’s key achievements in securing approvals for Remote and Shielded Operations. • Real-World Success: Learn how leading utilities have transformed their inspection programs with Skydio’s support.

+ Read More
TRANSCRIPT

Alright. Welcome to, the the last session here at Ascend.

It's great to have you all with us, and, I'm gonna be running through the drone program regulations and the Skydio approach to enabling Beyond Visual Line of Sight ops for a variety of use cases.

So to start with, I am Daniel Jenkins, regulatory program manager here at Skydio.

I'm part of the regulatory team with Jakey and Jen.

And our team is rather small compared to the rest of Skydio.

But we are really, really highly specialized in enabling your drone programs to have a quick clear route through FAA regulations and really enabling you to build both scale, repeatability, and reliability with your drone operations.

So we have a very lengthy track record.

Personally, I have worked in the Beyond Visual Line of Space Beyond Visual Line of Sight space for the past eight years.

I started at a detect and avoid company working with cameras on drones that could see other aircraft to avoid them.

From there, I went to the MITRE Corporation where I worked with local, state, and federal agencies in both designing their drone programs, developing their operations, and obtaining beyond visual line of sight approvals for those operations.

At Skydio, I've been here a little over a year and a half, and I've worked with over a dozen different customers in enabling, you know, both new leading types of operational approvals, as well as ensuring that, you know, folks that just need the standard kind of BVLOS route are able to get that and are able to get that easily.

So in the past year, we have had five different we call them precedent setting approvals.

Basically, that means it's the first time that somebody has done a certain type of operation or used a certain safety case, and we've helped pioneer those ideas with the FAA and that customer, ultimately resulting in a successful approval.

Beyond that, we have twenty eight unique regulatory approvals across customers, and then this is covered over thousands of sites.

So you'll hear me use a couple terms as we're walking through this presentation.

One of those is site specific.

Typically, that's just a handful of sites for a proof of concept style project, where, you know, you're limiting the initial outlay.

You're able to take this equipment and start integrating it into your workflows. You have to see how the drones work, the data they're collecting in a controlled environment. And then from there, we move into what we call performance based, which is how we can cover, you know, thousands of sites. These aren't thousands of different approvals per site. It's one approval that covers a large variety of sites. So for utility customers, this could be every substation across their service area.

This could be every bridge across the transportation network, etcetera.

Oh, there we go.

So the way that we typically think about the progression of drone operations is using this arc of autonomy. You know, it's kind of the sliding scale for how much human involvement is required moving towards stage five where it is largely automated and functions more as a, data supply where you have an idea of the data that you need. Say it's, you know, all components for transformers across all of your substations, you would then ask the system to generate that dataset.

The drones would go out and collect that data autonomously, deliver that back to your system. And instead of needing to plan a large effort where you have multiple crews going to multiple different sites to be able to collect this data, this drone systems are able to do this autonomously.

And so you interact with this system as more of a data delivery service than as a drone that you're going out and piloting.

So right now, we are largely in stage three, and that's what dock and remote operations really enable is instead of sending a human out on-site to collect this data that you see in stage one and stage two, we're pulling that human off the site, putting the dock on-site, and then they're able to connect in remotely and access that sensor to collect the data that they need. It's then sent back over the Internet to them. So without ever leaving their office, they're able to collect real time data at a substation or at a bridge or at a construction project, get that data delivered back to them, and make near real time decisions using that data.

One thing that, we demoed during the keynote was the multi mission, so one to many.

This can be multiple drones at a single site. So for, like, a large construction project or a large power generation facility, you can have multiple drones flying at the same time within stage four, collecting all of that data simultaneously, and then getting it back to you drastically reducing the amount of time it takes to collect that data.

Or you can have different systems at different sites that one pilot is then monitoring while they're collecting that data. So this could be half a dozen substations.

The pilot would then need to inspect.

Instead of going to each of those sites, driving times between each of those sites, collecting the data, and then bringing it back. They're able to connect to the docks, fly all those missions simultaneously, get that information back, massively reducing the amount of time, the amount of resources that you'd need to commit to get that data, and vastly improving the efficiency of inspecting across your, you know, networks or your different sites that you're working with.

So now we will dig into the waiver side. So this is kinda where the rubber meets the road. This is what we on the regulatory team do, you know, to enable these types of operations for you all.

So to start with, this one is pretty straightforward.

The pilot is out there. This is that stage one stage two type of operation.

They are able to fly up to four hundred feet AGL.

And this approximately doubles the distance that they are able to operate over standard part one zero seven visual line of sight.

So under this, the pilot is responsible for scanning that airspace, but they don't have to maintain line of sight with the drone.

So that's the big difference here between standard part one zero seven and then a b v loss waiver with that operator on-site.

One step past that is remote operations without a dock.

So, big utility use case here is for inspection of lines, potentially after an event. If there's a report of a downed line, you already know that you need to send a crew out. That crew can have the drone with them, and then somebody back in the office can connect to that drone, fly it remotely, get the data back that they need.

And as part of inspections, you know, this uses that drone to then get, the imagery into your workflows, your data pipelines, where you're managing these assets.

And since you already have the crew out in field anyways, you're able to use that person as a visual observer. So a lot like the previous one, that visual observer is scanning the airspace.

This increases the range that the drone can be formed from them out to about two miles, and then this supports up to four hundred feet operations without any additional waiver.

So we're gonna spend a little bit of time on this slide because, as you can see on the right, there are a lot of different variants when it comes to remote operations with a dock.

So here, you don't need to send anyone out into the field. You have that docked drone out at the site ready to go.

And this takes advantage of shielding. So shielding basically means flying close to things that crude aircraft don't wanna be flying close to.

So with substations, utility lines, construction areas, construction equipment, so cranes, bridge infrastructure, etcetera.

Typically, if an aircraft is interacting with your drone at this altitude, they have many, many other problems on top of the drone.

So we use this in combination with ADSB.

ADSB is automatic dependent surveillance broadcast.

It's a system onboard aircraft that basically shoots out a message every so often that says, this is where I'm at. This is my speed.

This is my vertical speed, etcetera.

So you can understand where other aircraft in the airspace are and avoid them if necessary.

Now to walk through some of these different types of shielding, so starting with the fifty foot shielded operations, this is kind of the easy button, the general case.

It's not tied to any specific type of infrastructure, and anything really that sticks out of the ground counts as shielding.

So trees, power lines, poles, buildings, you know, warehouses, etcetera.

As long as you're within fifty feet of one of those structures, you are in that shielded environment.

Now as we move to the right, you can see that two hundred foot, And this is a specific use case, so two hundred feet or up to the height of any object above two hundred feet.

So if you have a four hundred foot tall tower, you can go right up to the top of that tower and be within that shielded volume.

Or if you have a fifty foot tower, you can go up to two hundred feet over the top of that because you're still within two hundred feet of the ground.

This is typically used in locations where there's controlled airspace or something like that. Some other caveat where we wouldn't want an excessive amount of airspace over the top of the structure or in the event where you don't necessarily need to fly over the tops of these structures.

This enables you to inspect and gain additional altitude. So for mapping missions, things like that, you're able to do those more efficiently because at a higher altitude, you have a larger footprint with that sensor.

So we can also combine these two. So that's what you see in the bottom left, the substation plus utility line example.

This means that within that substation, within that critical infrastructure environment, you're able to go up to two hundred feet. And then the lines coming in and out of the substation, you're able to use that fifty foot general shielding over the top of those, which means that one dock within a substation can then inspect two miles of lines running in and out of the substation.

And for areas where that substation density is higher, Typically, this means that a dock in a substation can then inspect the line all the way down to the next substation. So if you daisy chain those docks, suddenly you have automated inspections for your entire network.

We also use the two hundred feet, over the top is, how we typically refer to it. So this is within those critical infrastructure sites. So this would be, you know, more like a power generation facility or a really large structure where, you know, you need that two hundred feet around it to be able to inspect, to have enough room to get all the data that you need. But you also need to be able to get over the tops of these structures and inspect them at a distance.

And what this one does is, you know, more for, like, uncontrolled airspace, areas where there aren't a large number of low flying aircraft.

This allows just more efficient operations.

It's something that the FAA is, you know, able to approve and enables you to have the most room to operate at the site and can support, you know, other use cases besides inspection. So mapping, analysis of how the site is changing over time, environmental surveys, things like that.

So part of what we do to enable those different types of operations is we build a safety case both with you and any, you know, unique characteristics that may exist in the environment that you're operating in, as well as our experience in dealing with the FAA.

So, really, the two core pieces here are the air risk mitigation and the ground risk mitigation. And if you think about a drone flying, there isn't really another category that the drone could cause damage while it's operating other than what it's flying around and what's below it.

So when we're thinking about the about the air risk mitigations, the big thing here is how do we avoid other aircraft, and how do we do that if the drone is not functioning properly?

Same thing on the ground.

How do we avoid running into things or people on the ground both during nominal operations when everything's working well and if components are going to fail on that system.

So we're gonna start with the air risk mitigations.

And the goal of this is really to maintain well clear and avoid what's called n max. So n max is a near mid air collision that's typically within five hundred feet of another aircraft laterally and then, a hundred foot vertical offset.

So the definition for Well Clear didn't really exist about six plus years ago, and there was a whole lot of math and a whole lot of work across various standards bodies by the FAA that ultimately results in this two thousand by two hundred and fifty foot volume or hockey puck around the aircraft.

And safe state is another kind of term of art here where when you're in a safe state, you're well clear by default. So if you're twenty feet off the ground and a plane hits you, chances are they're hitting the ground almost immediately.

If you're within five feet of a building and a plane hits you, they've already hit something else.

So you can't get further away than the safe state. Does that all make sense?

Perfect.

So when we're looking kind of across airspace, what you can see on the right here are returns for traffic flying around at different altitudes. So you can see one thousand feet on the top right and then two hundred feet on the bottom right.

And you'll notice there's a lot of cross patterns. As you get lower, those are runways.

So when we're operating near airports, which is, you know, also controlled airspace in some cases, there is an extremely low risk of encountering another aircraft, because pilots don't like flying low the entire time they're flying. You know, altitude is life, and altitude gives you time to make decisions in an aircraft.

So folks that are flying low, typically landing and taking off. So if you're away from airports, that risk just if you did nothing else, but were away from airports, that risk is very low.

However, the FAA, you know, typically wants that risk to be as close to zero as possible.

And so there are some, you know, mitigations that we put on top of just operating away from airports. So, generally, the low risk category of operations is gonna be greater than three miles from airports and greater than half mile from heliports.

When you get inside of that, that risk starts to go up the closer you get into the airport.

And if any of you are familiar with the Lance grids or the facility maps, as you get closer into the airports, those altitudes decrease.

Often, it's set up like a upside down wedding cake where the highest risk right around the airport is gonna be zero feet or maybe fifty feet. And then as you move out, that altitude steps up as you get further and further away.

So one way that the FAA has, agreed that there are some, you know, risk mitigations for the air risk side is with noncooperative detect and avoid sensors.

So at the top there, that was my first job, flying drones at planes out in the desert and testing onboard collision avoidance.

Moving down, there's a little radar there. A bunch of colleagues from my first job ended up moving into that company. And then while I was at MITRE, I assessed various DAA systems.

So acoustic systems that are listening for aircraft, the electro optical that are using cameras to see it, and radar, which, you know, radar is a little bit more ubiquitous.

Each of these systems has different limitations and different restrictions on what they're able to see, how they're able to be used.

They require setup. Typically, they require power. And so as you're integrating these types of systems into your operations, it gets more and more complex.

Additionally, you have to be able to demonstrate to the FAA that the system works and works well enough to count as a risk mitigation for your operation.

So I've talked about the DAA performance standard a couple slides ago.

There's a whole series of tests and test test methods and metrics that you need to satisfy for the FAA to approve a detection void sensor.

Additionally, they have limited range, and so if you're looking to fly ten plus miles, you're either gonna need a really expensive radar or you're going to need multiple sets of these systems set up along that flight path.

Operationally, they can also be pretty challenging to use.

So the bottom right there is definitely a little bit of a throwback with mission planner.

The way you can see there, that pink line is a flight path, and all of those green points are potential loiter points.

So if the system triggered that there was an aircraft in the area, then the plane is gonna fly to one of those points and loiter.

And with all of them stacked up on top of each other for longer and longer missions, that's more and more different things that that pilot has to keep track of. And you start getting into information overload, and it's more about managing the DAA system that it is flying the mission and getting the data that you need.

So this is why we use shielding because you don't need to prove that the ground is the ground. You don't need to prove that planes flying into the side of a building or the side of a mountain is bad and that they don't want to do that.

That's all very much understood.

And the shielding altitudes are agreed upon based on, you know, some research that was going on at the same time as the DAA research was going on.

And so we try to default to the least friction solution to get you the data that you need.

So ADSB, as I mentioned, is one of those tools that we use on top of shielding.

So with the FAA rules around ADS B equipage, they've generated what we refer to as ADS B out airspace. So around the larger airports where you have that mode c veil, which is a thirty nautical mile ring around the airport where there are certain reporting requirements that the FAA places on all the aircraft that enter that airspace.

ADS B is now included.

And, occasionally, an operator can get permission to turn that off, but that's something that the FAA has to approve on a case by case basis.

Given this and that operators are typically flying in, near, or around one of these airports, we can use this increase in the amount of aircraft equipped with ADS B to further help that safety case that ADS B plus shielding is going to limit the potential for those air risks to an acceptable level for these operations.

So with that ADS B information, we are able to and we've kind of pulled the map off the back of this image here. But what you can see is a blue PIP pointed the direction that your drone would be flying. So in this scenario, north.

And then there's a red PIP here with some additional information beside it.

So these items are transmitted from that aircraft, and we are able to get collect that information from a receiver at the dock and display that to the pilot.

So here we can see that the, you know, intruding aircraft here in red is two hundred feet above the drone, flying at a hundred knots, and is climbing at fifty feet per minute, and is point four five nautical miles away.

Now you see the alert here at the top, and this is a little bit more dynamic, but this alert pops up when that aircraft is closer than the avoidance time line or at that avoidance time line distance.

The avoidance timeline comes from FAA research in human reaction times, you know, how quickly you can see something, understand what it is, and then start acting to resolve the collision.

We also include the time for the drone to get to the safe state. And so when that alert pops up, that means you have enough time to identify it, command the drone to descend, or get close to a structure.

And then while you're in that safe state, you're well clear. You can wait for the aircraft to leave and then resume your mission.

So for two hundred foot operations, it's about thirty three seconds.

For fifty foot operations, it's about fifteen seconds.

And an aircraft with traveling at the standard hundred and thirty knots is gonna cover about one point two nautical miles in that timeline for two hundred feet and about half a nautical mile for fifty feet.

This technology is lightweight, easy to use, integrates directly into the GCS, and doesn't directly command the aircraft, doesn't interfere with the operation.

It lets you know if there's a risk, and then the pilot can immediately respond to that risk, wait until it's cleared, resume the mission.

So this is the big advantage of, you you know, using ADS B versus detect and avoid systems.

Oh, let's see.

Wanna advance we might have died here. Alright. This is the the fun one with even more clicking.

So this is gonna be, an interactive portion here.

The point of this is to show some of the decision making that a pilot is going to be making when interacting with ADS B.

So a low flying aircraft that is within that one point two nautical miles doesn't necessarily mean that you have to avoid. So remember that two thousand foot by two hundred and fifty foot hockey puck? That's what you're trying to stay out of.

And so let's click the first one. Okay. So what we can see here is that we have an intruder going to the bottom left.

They are two hundred feet above the drone.

They're not changing their vertical vertical altitude at all. So they're zero feet, per minute on that vertical speed. And they're flying in a hundred and thirty knots. They're one point two nautical miles away. So do we think that the pilot would need to avoid if they see this on their screen?

Hands up if yes.

Perfect. Alright.

Let's click it. So this one's good.

They're traveling away from us. We're traveling to the north.

They are greater than two thousand feet horizontally.

And even though they're within that two hundred and fifty feet vertically, they're flying away from us. We're flying up.

There's not a reason to stop the mission because there's no risk here to avoid.

So let's hit the next one.

So with this one, it is flying straight toward us seven hundred feet above.

Vertical speed is zero, so they're not changing altitude. Flying at a hundred and thirty knots, one point two nautical miles away.

Do we think we need to avoid this one?

Yes? No?

Yes. Little mixed.

So with this one, we don't need to avoid right now.

But what we do need to do, because it is flying toward us, is make sure that that vertical speed doesn't go negative.

So they're seven hundred feet above us. If they start descending, then we need to immediately start descending.

Some organizations in this scenario may just recommend as a default.

Okay. If somebody's coming towards you, doesn't really matter the vertical separation at that range.

You need to get to the safe state, wait until they're passed, then you can resume the mission. That would be an abundance of caution, but generally speaking, no. You would not need to maneuver in order to remain well clear, and let's go with the last one.

So this is in two hundred feet of separation. Same thing on the vertical speed. One point two nautical miles away headed right toward the nose.

Do we think we need to avoid this one?

Perfect.

Alright. So, yes, with this one, you know, once they're within that two thousand feet horizontally, we will have breached well clear.

At one point two nautical miles away, this is the range where at two hundred feet, we have enough time to get to that safe state before we've breached well clear, and so we would engage that safe state maneuver.

So part of the training with using Skydio systems, using the dock with that ADS B is understanding how all of this works. And then part of pilot training, you know, both under the waiver and generally is going to include that safe state maneuver. So they're gonna know, okay. I see this. I hit this button. I wait. Make sure the drone is coming down.

When we're in the safe space, we're good to go.

So that is all the air risk side. So when we're looking at ground risk, you know, how do we keep the drone from injuring people on the ground, both when things are working and when they're not?

So there are a couple different strategies here. One is to fly out in the middle of nowhere where there aren't any people, then you don't have to worry about it.

When you do get into areas that are more populated, and this is a project that we worked on with a school, two years ago.

There are a couple strategies kind of tactically and strategically that can be used. So strategically is flying in areas where there are people are not going to be. So in this image, flying over the roofs of buildings, flying around that forested fence line, and, generally, you know, not overflying parking lots, not over overflying those crosswalks.

And then tactically, so this is day to day, you know, minute to minute, different things that you can do.

In this scenario, we know that classes are going to end and start at certain times. We know that when we're between classes, more people are gonna be outside moving between buildings, and so during class transition, you don't fly.

Once everybody's in their classes, then you can fly your mission.

You can also use the camera on the aircraft to effectively look both ways before you cross the street. It's a really simple strategy, but it allows the pilot to make sure that before they would fly, you know, from building to building, they can look, make sure there's nobody between them, and then fly across.

By doing that, even though there is the potential for people to be there, you have that check before you would move the aircraft in order to ensure that you're not actually flying over those people.

So next slide.

So with those pieces together, you have a repeatable and reliable Beyond Visual Line of Sight approval because those are all the boxes that the FAA is looking to check. Have you adequately mitigated air risk? Have you adequately mitigated ground risk?

Do you understand the system that you're operating well enough to safely operate it? If you can, you know, convince the FAA yes for all of those items, you have a pathway for those Beyond Visual Line of Sight approvals.

Next slide.

So as far as the regulatory side, what do we have upcoming now that you've seen all the cool new technology with the new dock and the X ten?

So for those of you that have used the x two x two dock, have done those deployments, you know that on top of the dock and on top of the aircraft, there's a lot of other equipment that we bring out and integrate with those systems.

So what you can see in the top left there is the weather sensor, and then that, white antenna is the ADS B.

With the x ten dock, we've integrated those into the dock itself.

So instead of running power to those, power to the dock, you just run power to the dock. It's all integrated in there.

And we have also increased some of the redundancies with this system.

So the access points, that serves as both, external radio, so you have a different line of sight to that system. So if you're flying around, you know, more congested area, you have multiple external radios out there, you can ensure that you always have line of sight to a radio when you're flying using that.

And that external radio will function even if one of the other radios that you're using fails, and so you have a fallback option.

If any of you have, you know, worked around or, you know, seen, videos on commercial aircraft, you know that redundancy is a key component to those systems.

By integrating redundancy into your drone operations, you increase that level of safety because, you know, like I was saying earlier, you have to prove that the system works both nominal conditions when everything is working and off nominal when there are components that fail. If you have fallback components, that supports that safety case.

Excuse me.

So by reducing the number of different components that we're connecting to everything and simplifying the process of getting these systems out into the field, it makes it a lot more effective and efficient to deploy these systems across your areas.

And so with all of that, I know we covered a lot of ground here today, but I wanna open it up for questions, that you all may have, and then we'll have a mic coming around.

So this can be questions about anything I went over or more general, you know, drone regulatory stuff.

Oh, right here.

This is the actual question. It's easy.

Do you have a a page on your website that goes into this in more detail if we have questions, and wanna learn more about this issue?

So this I know we're recording this, so this presentation will be available.

As far as the information, we do have some on our website, but we're also happy to sit down, you know, discuss, like, a specific use case with you, specific, you know, integrations with different systems, and, make sure that we can tailor whatever we're going to be doing to what would best work for you.

Uh-oh. No.

We kind of was mentioning earlier at lunch, how existing BBOS waivers, might change or be impacted, you know, with the doc solution.

Just to kinda circle back to that Mhmm. Can you, explain, like, what what the dock introduction of the dock, how that's changed, how the FAA sees VWOS waivers?

Like, I guess, with new docs coming on scene, is there any type of did it make it easier or harder for to get VWOS waivers when you introduced the doc?

Yeah. So the doc enables slightly different modes of operations. So without that dock, the drone has to get out to the site somehow. Typically, that's a person.

So, like I was talking about with the non dock remote ops where you have that crew taking the drone out and then somebody connects remotely, what the dock does is remove the need for that person to go out.

On the regulatory side, what that means is you can't use a visual observer anymore unless you're sending a person out to the dock, which kind of defeats the purpose of the dock.

So with the dock, shielding and ADSB are the primary risk mitigations that we use. So operationally, that's the difference. Instead of a VO and four hundred feet, for that altitude, it's going to be shielded with ADS B. So either fifty feet or two hundred feet. So that's the big operational difference.

As far as the equipment that goes out, when you have the dock out there, that's also handling the data transmission, you know, getting whatever you've collected off the drone, piping it back to cloud.

When you're using a person, then that person needs to pull that data off.

Now as far as the approvals, so FAA approvals for Beyond Visual line of sight generally are system specific. So if you say, I'm going to fly a Skydio x two in your waiver to be able to fly an x ten under that waiver, you would need to amend. And, basically, what the FAA needs to understand is the new equipment. So you need to provide information about whatever new system you're gonna be using.

And with that information, there are a few different ways, that you can approach it. You can just say, you know, I want to fly the x ten and the x two. I wanna change nothing else about my prior approval. I just wanna use this additional system.

You describe that system, and then the FAA can, you know, take that information as long as that new system meets the risk mitigations and requirements that you had under your previous waiver. They can update that waiver, and it will supersede the previous one and then cover you with that new system.

If you want to do something new in addition to what was in the prior waiver, it's going to be a new submission and you would include all the systems you wanna use, the new types of operations.

If you've already submitted a waiver in the past, you can use a lot of the the information that was in that application. You know, anything that's changed, you can add. Anything that's the same, you leave. So it reduces the amount of work that you need to put into that new application, but the FAA would then take that and approve it on its own. So it would be a a separate approval from the previous one.

So you're you're gonna have that first approval. That's gonna be for that, you know, initial system and that initial operational set. And then with that new approval, you would have new systems, the new operations that you would wanna do, and very likely, you know, the the stuff that you had under the prior waiver. So you'd be able to operate under either.

Typically, you know, companies are gonna shift forward to that new waiver because keeping everything tied to to one process, one workflow, and one approval number is gonna be, you know, the most straightforward solution.

So so, ultimately, do you think that a is gonna make it easier to if you have a doc, are they gonna do you see the trend where they're like, okay. You have a doc. It's not of a longer review, or do you see a doc complicating the review process?

It's kinda what I was yeah.

No change as far as, like, complexity of the review process or the review timeline.

The FAA by now is very familiar with Skydio and Skydio docs.

We've been working with them across the handful of approvals for x ten doc as well to support, you know, both the demos that we did here today, initial testing of the system at headquarters, and a couple upper other opportunities that we have, you know, across the states to make sure that this system functions in different types of environments.

So we've been pursuing those approvals already. The FAA has accepted them.

And so when we come with customers describing, you know, especially operations that have already previously been waived, that means the FAA understands the safety case. They understand the operation and the risks. And if they've already approved it once, the odds of them approving it again are very high. And so on the back end, like, another part of my job is communicating with the FAA, helping them understand the technology, understanding what's new, what's different, how it functions, how it performs, and making sure that by the time, you know, customers are coming through this process, the FAA is already very familiar with how the system works, the risks, how we're mitigating any of those risks, and enabling these operations.

So the timeline is gonna be that sixty to ninety days. That's kind of a default for the FAA. Sometimes it's a little shorter. Sometimes it's a little longer.

But as far as, like, the complexity I was describing with the detect and void systems, we don't have to prove that the doc works in every single specific environment with a whole test campaign and all of the things that would go into those systems.

With the dock, it's not flying. So the FAA cares a little bit less about it versus the drone that comes out of it. So when the drone is flying, as long as that dock is able to effectively communicate with the drone and control it, that's really the part that the FAA cares the most about.

So That that's good. Thanks. Alright. Appreciate it. Yeah. No problem.

Two questions. One on the ADS B in, is that active now on the x tens?

Do you mean are the x tens broadcasting ADS B?

No. Are they receiving? And is that a active function on the x tens, or is that something to be turned on later?

That will be turned on later. Okay. So we're still, you know, independent of, you know, the the a d s b antenna with the x two dock. You know, there's there's still testing that we're doing with that system.

But, you know, when it when it is ready, it'll get turned on.

Okay. And then second question on the the shielding, has the FAA ever asked for, like, verification that you're maintaining fifty feet or two hundred feet?

No. So there are a couple different strategies for, you know, procedurally ensuring that you're remaining at that altitude or maintaining that separation.

But as far as the FAA wanting, like, flight logs of your exact positions and basically demonstrating it or the FAA proving that you weren't is not gonna be a factor unless something happens.

Now if an FAA person, you know, visually sees and you know they want to come and ramp check you effectively, they absolutely can do that.

And if you're routinely flying outside of that shielded volume, they with as part of the waiver, they can audit you at any point in time. So they reserve the right to basically tell you to turn everything over so that they can review.

So when you're operating, you wanna do your absolute best to make sure you're maintaining that shielded volume.

But there's not, like, an automatic flag that'll pop out to the FAA that says, oh, you popped out of your shielded volume.

Yeah. I was just thinking more in terms of a method that you have to include in the waiver, application.

So the the method that we use, typically, is going to be, you know, you have the altitude that the drone is at. You make sure that's below whatever your shielded volume is.

When you're navigating to structures, you know, for, like, a large warehouse, You fly up to the top. That's gonna tell you a number.

You add whatever your shielded volume is to that number, and then now you know the maximum that you can fly relative to that building.

Thanks.

Staying on the kinda safety case topic, have you guys found a good way? I know you're doing mostly shielded operations, but are you found a good way to do ground risk assessments with population density, traffic counts, things like that?

When I was back at MITRE, I did a lot of that for the, Massachusetts Department of Transportation.

A good way? No.

With ArcGIS, and, it was the LandScan dataset.

There were some strategies to, do various summaries of, like, the, population density within the grids and then, you know, taking a flight path, overlaying that, and then being able to identify density within each of the grids or across the site, that gets very complex.

Happy to talk about it with you afterwards.

But as far as, like, needing to prove for these b v loss operations that we've done, we haven't needed to prove a specific ground density, or things like that because really the fundamental rule right now is that you just do not fly over people.

And so we use those strategic and tactical mitigations of visually checking, flying over areas where there won't be people, to mitigate that portion of the risk versus making the ground density and exposure style arguments.

Just wondering, in your opinion, what is the path and challenges to a fully autonomous FAA waiver?

I am working towards that in in some respects.

You know, we have a very small team, and we're cranking out a lot of waivers. So if there was a button that I could just hit, I would be very happy.

That said, there's a lot of very specific language and structuring to these waivers, and it's a little bit of art and a little bit of science.

So there are some strategies with, you know, like, find and replace style. You can do Excel integrations.

AI is kind of starting to make its way over, but AI has its own challenges with making things up. And when you make things up with the FAA, they're not a huge fan.

So right now, one of the big benefits with, you know, using the same aircraft across all of our waivers, you know, I'm not writing for other aircraft manufacturers. I'm writing for Skydio.

So a lot of things I can keep constant.

And then with the different flavors of waiver, you know, I can have some base templates, based on the different mix and match, but every single one of them has custom work that goes into it based on what the customer needs, based on any unique portions of the operation that we need to describe and make sure the FAA understands how we're mitigating.

So one zero eight is gonna happen well before any sort of, like, end to end automation of the current waivers.

So if they, you know, if I were able to make that today, I'd be able to use it for, you know, a couple years, and then one zero eight comes out, and I'd have to scrap it and start everything over with the new rule set. So there's definitely some aids for generating that, but as far as end to end fully automated, there's just too much custom work and a little bit too much variability for that right now.

I think my question was more autonomous flight.

Oh, oh, so, like, supervised, fully fully supervised?

You know, no no humans, like, you know, fully supervised. You know?

So there's some progress that's being made in that direction.

The big thing right now is controlling that operation. So this is for, you know, specific sites with specific criteria, fences so we can make sure that people aren't walking through, different things like that. So they are taking some baby steps in that direction, but we don't have one that, like, a general customer could use at the moment.

Yeah. So I was just a little unclear and wanted to see, like, if you foresee in the future. With the addition of ADS B in, you had mentioned like, you get the alert on the screen of the controller.

Is there gonna be are you gonna take the decision making out of the human, right, to decide, okay, this is coming from a certain direction, certain altitude.

Will the drone be able to get to that safe volume of space on its own if it calculates that it's necessary? Or is that do you think that's always gonna be, you know, the responsibility of the pilot in command?

So, ultimately, no matter what, it is the responsibility of the pilot in command as far as the FAA is concerned.

Right now, keeping it with that pilot giving the command is something that's a lot easier for the FAA and understand.

And operationally, drone pilots don't always like when the drone does something that they haven't told it to do.

And so this keeps the most people happy.

As soon as automation of the control of flight starts happening, beyond, you know, what the pilot expects or beyond what they you know, when you start, like, a three d scan or something, you're clicking, like, yes. Okay. This is how I want it set up, and then it does its thing.

With ADS B, it would just be flying along, and then boom, does its own thing while you're looking at the sensor feed.

So right now, no plan that I'm aware of.

But in the future as this automation starts increasing, like, I'm sure that's gonna be, something that we're about to integrate. And then once it gets to the, you know, level of fully autonomous, then it's gonna have to exist because there is no pilot to then command the ASB ASB maneuver.

Maybe a stupid question, but once we get the BB loss waivers for any class g, class c, aerospace type of, complex environment. Do we for every mission we do, do we need to report back to FAA for anything once we have the waiver obtained?

So not typically.

As part of the certificate of waiver, there is a specific set of record keeping requirements, that the FAA will tell you. So most of the time, the responsible person has to track, maintenance for any aircraft flown under the waiver. They also need to track any pilots flying under the waiver and the training for those pilots.

But as far as mission by mission, there's not a requirement.

One thing that you do need to be aware of is that they do specify the different types of maintenance that qualify that need to be recorded.

And there's also something called a functional check flight, which basically means just taking the drone out under regular part one zero seven rules after any of those maintenance items, flying it, making sure the maintenance was done properly, making sure everything works before you then go fly it beyond visual line of sight.

So that is typically the most onerous, like, record keeping component is maintenance and functional check flight logging.

But as far as flight by flight, no requirement.

And then the the FAA reserves the right to kind of audit any part of the program when you're operating under a BB loss waiver.

Alright.

Well, thank you all very much for coming out, and, I'll be here for a little bit longer if you have any other questions.

And, hope you all had a great Skydio Ascend, and, we'll see you.

+ Read More

Watch More

52:00
Navigating Regulations for DFR
Posted Sep 27, 2024 | Views 418
# Regulatory
# Remote Operations
# DFR
# First Responders
# Drone
# Patrol-Led DFR
# Public Safety
Skydio Dock for X10 Keynote Presentation
Posted Sep 27, 2024 | Views 10.9K
# Skydio Dock
Substation Inspection from an Office Chair
Posted Mar 23, 2024 | Views 7.7K
# Inspection
# Skydio Dock
# Utilities
# Remote Operations
# Maintenance
# Drone