#80 Joerg Dittrich, JARUS & DLR: Drones, risk, and SORA 3.0

Joerg: 0:00

So what's the prerequisite for actually using UTM? You need traffic. Because it manages traffic. When you don't have any traffic over a particular area, you don't necessarily need a UTM system. Think of UTM like a traffic light at an intersection. Are traffic lights bad? No, they're fantastic. They're a great solution to a problem. Do you need a traffic light on every intersection of two country roads? Not really. There's other ways of mitigating that risk. I get the impression in some of the debates over UTM or U-Space in Europe, not generally speaking, but sometimes I get the impression like, we're going to install this new U-Space or UTM system in that remote area, and when we install this, traffic will come. The analogy would be, we have a country road intersection, we're going to build a traffic light, and then we know that more cars will come. That's not going to happen.

Luka: 1:54

Today, we're thrilled to welcome Joerg Dittrich, a leading authority on unmanned aircraft and drone regulations. Joerg serves as the leader of the Working Group on Safety Risk Management for the Joint Authorities for Rulemaking on Unmanned Systems, also known as JARUS.

2:11

JARUS

Luka: 2:11

is best known for creating the specific drone category and the SORA the Specific Operational Risk Assessment methodology, which has been adopted by numerous countries around the world, including those under EASA. As the head of this working group, Joerg oversees the creation of safety and risk management guidelines for drone operations, with the SORA framework being one of the most significant contributions. Joerg also brings 22 years of experience at the German Aerospace Center, DLR, where he's a senior expert in drone regulations and a scientific advisor to various national and international task forces on drone rulemaking and compliance. At DLR, he has led numerous UAS projects focusing on areas such as unmanned helicopters, manned- unmanned teaming, automated mission planning and real-time image processing for flight control among others. This episode has been in the works for over a year, as we eagerly awaited the release of SORA 2.5. And now that this milestone has been achieved earlier in the summer, we sat down finally with Joerg to discuss the current state of the drone industry and the regulatory landscape that enables complex operations. Our conversation covers a range of topics, including the evolution of airspace integration for drones. Namely before the mid 2010s, the expectation was that certified unmanned aircraft operating at high altitudes under instrument flight roles would lead the way and be integrated into the airspace first. However, the rise of hobby drones in the 2010s shifted the focus to very low altitude operations, spurring the creation of regulations like the specific drone category and the SORA framework. A major focus of our discussion is the specific drone category itself, which represents an odd concept compared to a traditional aviation. Unlike certified traditional aircraft, which need to gracefully remain in flight, this category emphasizes an acceptable risk approach. Meaning, crashing is okay, as long as it's crashing gracefully. Joerg sheds light on how risk differs from hazard, the safety targets for third parties, both on the ground and in the air, and the methodology for assessing and mitigating air risk. He even gives us a glimpse into what SORA 3.0 might hold for air risk mitigation. We also dive into topics like drone delivery, agricultural drone applications, UTM, U-Space, Part 108, and the transformative potential of drones in various sectors. So enjoy this engaging conversation with Joerg Dittrich. Joerg, welcome to The Vertical Space. It's great to have you on.

Joerg: 5:36

Well, thanks for having me.

Luka: 5:37

First question, one thing that few in the industry agree with you on.

Joerg: 5:41

That's the only question I did expect because, you know, I've been a regular listener.

Luka: 5:45

Okay, great. So what is it?

Joerg: 5:48

So, recently there has been a lot of talk about, urban air mobility, and now it's called, Innovative Air Mobility, or there's probably even more acronyms around that. And there's a lot of talk about air taxis, and, if you go to any drone event, you will see, either mock ups or, videos or flight tests of these awesome machines. And there is basically a lot of focus about these kind of new entrants into the market. And, since you've asked me, where do I disagree? I believe, my gut feeling is that most of these, operations probably require, pilots in the first phases and with related transition to operation with the no pilots. I believe what will happen is that this is an application that is very interesting that would find its niche. However, I do not believe that this will alter people's lives like everybody's will barely be affected. There will be some operations that will be affected. But the big thing that I see coming is the big growth will be in unmanned cargo applications and mostly unmanned applications. They have the potential to actually change every day's life. I just recently managed to, see, a big drone operation in, Ireland where they're delivering, food items to customers in one on suburb. And people love it. People order it all the time. And this, I believe, when this is going to scale up, this will be a big, big business. And it will literally change the way people live their lives, not only in big cities, but also in remote areas where certain services suddenly become accessible. We can already see that in some applications for example, Zipline is doing in Africa. That's a great example. There's very, a lot of other similar examples. I believe we're still in the infancy of that type of industry. It will become so big that our children will find a lot of things normal that we would basically say, that's not going to happen.

Luka: 7:40

Why do you think that this is a contrarian view? That seems pretty straightforward, at least, within the four walls of, our organization.

Joerg: 7:49

Well, let me give you a little story here. Just a couple of weeks ago, I was honored to be a delegate at the ICAO AAM symposium. That was the first time ICAO hosted something on the topic of AAM. And it turned out it was the biggest event ever hosted in the 80 year history of ICAO. They had, I think, 1, 400 attendees, even more people registered to attend. They had to open overflow rooms during their presentations. The ICAO Council Chamber was filled to the last seat, if you've ever been there. It's huge. So there was clearly something that has changed. I've been attending meetings at ICAO for years, but this time it was way bigger. And the reason it was way bigger, because everyone is talking about, passenger operations with, drones, about all these new configurations. There was exhibitions, there was mock ups, there was all these great things. And so the focus was very obvious on these kind of, new entrants. At the same time, I'm saying, yes, they will exist, they will have their place. But I believe that the business on the unmanned end is bigger. That's my gut feeling. Some of the, in the industry already agree with me, but your question was specifically what most people disagree with. And if I look at the reality of what happened at these shows is there's so much attention being directed at these type of operations and so little at, urban cargo, remote urban car, or remote, area cargo and these type of operations.

Jim: 9:12

If you had gotten in front of the ICAO group at that time, most of them probably would have disagreed with your claim, probably not everyone, obviously.

Luka: 9:20

What about on the drone side of things? Obviously you're deeply involved in, SORA, we'll get to that. And, the smaller UAS end of the market. What are some of the contrarians views there?

Joerg: 9:33

How you can integrate, into the, airspace. What I mean by this is, what we see today is, there is many regions of the world, that allow BVLOS operations. The FAA uses an exemption based process, we have, a process in place in Europe that we don't require exemptions. That being said, there's multiple ways to achieve the same, and often boils down around the question, well, when you're flying BVLOS, can you actually, see enough other traffic in the airspace that you're using, and what mitigations are in place to assure that your flight will be collision free? And that is something where there's still a lot of disagreements. There are some methods that are being accepted in some parts of the world, where other countries say, well, you kind of do this here. And we are not really harmonized looking at the, really at the global landscape.

Luka: 10:23

We know you as a very, opinionated, independent, thinker, and I know there's more, there's more to it than you're probably hinting at right now. So I want to tease that out of you somehow.

Jim: 10:35

Well, he just threw the entire advanced air mobility market under the bus. So that's something.

Luka: 10:39

Yeah, yeah, true, true.

Joerg: 10:41

I'm not throwing it under the bus, actually, because, again, as I said, I think, first of all, I am an aviation enthusiast, and some of the configurations that are being proposed are so awesome, I want to fly with it. So, you know, yes, I'm excited about it. The question is, how will the market develop as an overall market? Will it be mostly, big, passenger carrying drones? Or will it be more of the cargo carrying drone Where will we see the biggest, change when it comes to what affects our lives? The real, amazing thing about these kind of early operations and the similar what you have in the, in the Dallas area is these kinds of things under, if you set this up properly, there was often this long time debate about, you know, the population, the average population will only accept if it is for medical reasons, they have drone, have drone flights about their homes. And I believe, you know, we are getting closer and closer to establish the fact that this is possibly not true. Because, people tend to actually love drone delivery and they love it even more when it is for their benefit, not for somebody else's. So I mean, that, that's probably a paradigm change that we're going to observe. It was still in the early stages because, you know, let's face it, what percentage of the population has access to drone delivery service? That's like a fraction of a percent.

Jim: 11:59

I know we're gonna get into use cases in a few minutes, but that's a compelling vision you set forth. What, use cases would change people's lives that would be delivered by drones that you think may not be as obvious to some of our audience?

Joerg: 12:12

Well, um, one of the most compelling things, if you remember when you had, Bobby Healy of Manna on your, podcast, he mentioned that, you know, the best growing product is coffee. Um, it's a mystery to Bobby why it is coffee. It's an even bigger mystery to me. And what that teaches us is, if we are going to predict what the best use cases are, we most likely will be wrong.

Jim: 12:33

But you made the claim that our lives will be changed. Clearly the high volume coffee delivery is, it seems to be successful right now. But you, what you're saying is that the drone will, provide services in a way, they're not being provided today. Why are you so confident this is the path of the future?

Joerg: 12:52

The real question is, um, drones have been popular for a while, and I've been in the drone industry for more than two decades, and one of the big advantages of drones, usually, that's being mentioned is that they're cheaper. Well, actually, if you look closely, most drone operations are not actually cheaper. The real change will happen when they become cheaper. And with the first kind of drone operations seem to get so cheap that, you know, use cases, could be, exploited that, you know, require them to be affordable. So if it is affordable, if it is more affordable to send a drone than a courier or, you know, for food delivery or any other, uh, service, for that, that price point, if I can hop on on public transport or get in my car, drive to, uh, a particular hardware store because I need a particular item. And I look at what, what was it going to cost me effectively, you know, depreciation of the car. Um, uh, fuel cost, time spent not working, but driving, uh, to get this item that I need. And if the drone delivery is at the same price point or at a similar price point, hey, why wouldn't I go for it? I would say, yeah, you know, I'm not going to drive out there. And it is these kinds of things, like we cannot imagine how it's going to affect our lives. So I'm not really good at drafting what the future will look like. Go back a couple of years when nobody had a smartphone. The first smartphones came to the market, you know, the ones that were non Apple. People say, yeah, that's interesting, that's a technical marvel, but, you know, we don't need this. Then Apple, you know, brought up the iPhone, the rest is history. And, imagine you travel back to that time and tell your younger self, That, you know, you're going to agree to have a device with you every day, you will never forget it at home, it will track your position, it will do multiple updates to a couple of companies, uh, that you have no control over, and you'll be happy with it.

Luka: 14:50

And, and you'll be able to control your, sprinklers in the garden with your phone or, switch on and off your Wi Fi enabled lights in the apartment, that, I mean, I hear your point loud and clear.

Joerg: 15:04

Back then when these things came out, it was a novelty. Nobody would think what you could actually do, and there's so many applications that are very novel to us and for our children. Like my, uh, daughters grow up to be digital natives, and I consider myself a computer geek, but when I look at what they think is normal, uh, that's already different from what I think. So I believe, yes, These kind of operations have the potential to change our everyday life.

Luka: 15:31

In one of our earlier conversations we had a really interesting, discussion around airspace integration for drones, and how in Europe there seemed to have been a paradigm shift between expecting that this will happen top down as opposed to, you know, the bottom up that evolved. Do you mind elaborating on this? And giving the audience a bit more context as to the evolution of the thinking of integrating drones into the airspace.

Joerg: 15:58

Okay. yeah, sure. This actually doesn't apply to Europe. This is more like a global paradigm shift that I witnessed around the year, um, 2014, roughly. What we had before was that everyone in the drone business was pretty much convinced that the first drones to be integrated into the airspace that would be allowed for civilian operations would be high flying drones that fly at high altitudes, that follow instrument flight rules, that are certified aircraft. And this was obvious, and really the only little problem that needed to be solved is how do we successfully detect and avoid, so that it is acceptable, that we can certify such a detect and avoid system. And then this integration would happen, starting with IFR flying. And the big challenge would be VFR flying at low altitude. Well, you don't really know who's flying there. This is really difficult. And the paradigm shift is strongly connected with the, uh, emergence of the hobbyist photo drones, like the DJI drones, the Parrots. They would be used at very low airspaces and, you know, first regulation had to be developed to accommodate them and we had this 500 foot thing that emerged that like, you know, actually most manned aviation flies above 500 feet in most countries, in Europe, for instance, we have, single European rules of the air. It, basically prescribes that, you know, you shall not operate an aircraft, below 500 feet, so, unless you really have to for landing or, for, training purposes and so on. So, usually you fly above that. So, the airspace below 500 feet, the VLL, is more empty. And that's when this transition happened to why don't we start with these guys first. They don't have to fly IFR, we don't have to wait for all the ICAO annexes to be updated by the work of the ICAO RPAS panel. So, that's when this shift happened. And around the same time, around 2015, the drone community first heard of the idea of a specific drone category where you would first assess what the risk level of a drone operation would actually be, how hazardous it could potentially be, depending on type of airspace used, size of the drone, location, and so on. And then you could derive what kind of provisions you would have to ask for in order to authorize that kind of operation. So, when Europe decided to, adopt the concept of the specific category that was, uh, in early 2015, that was a big, shift. Europe certainly did not invent that alone. That was a, collaborative, development, but, uh, they were the first, major region that said, you know, actually we believe in this concept. We should make this a part of our rule system.

Luka: 18:40

So talk about that a little bit, because you've had a front row seat to Europe embracing this risk proportional perspective on certification and I recall another conversation with you where you kind of brought the example of, Hey, we have a certified category of air vehicles where crashing is totally not an option. And now you have this emergence of a, specific category of drones where it's okay to crash, but, you know, let's crash gracefully as opposed to gracefully staying in the air. That, sounds like a big shift for traditional aviation people to accept. What was that like in the early days of coming up with this idea.

Joerg: 19:26

Well, this idea was, as part of the development at JARUS where this concept of a specific drone category was, invented and, further developed. Um, it became obvious that, you know, if you want to assess the risk of a drone operation in order to provide a level playing field, you need to have a standardized way of assessing risk, which eventually became, the SORA. And the real issue, why this is very difficult to implement to begin with, is, uh, if you look at the way aviation authorities usually work, they usually have, separate departments. You have, the, airworthiness people that take care of, technical reliability of aircraft. You have the flight standards. On the flight standards, you find, licensing. You find the, authorization of organizations. And part of the idea of specific category operations is, which part of these three columns of aviation safety do we need to what extent, and how can we do trade offs between them, so to say. And the real issue this produces is that, the way aviation authorities are structured is not really supporting that. So it becomes, often becomes an organizational problem. And, but this is where a lot of reluctance, came from, in my opinion. Now, in Europe, it became, law. Not only in Europe. We have, uh, several countries outside of Europe that have very similar, rules. And, use very similar means of compliance, uh, that what we use in Europe. And, bottom line is, let's look at the drone operation first. What is the kind of operation? Is it, what are the typical characteristics of that operation? What hazards are involved? And based on that, you get assigned, a certain risk level. And, uh, then from then on, you know, you need to, you can develop your drone or you need to, what kind of drone to buy and so on.

Luka: 21:13

Okay, explain the SAIL levels, please, for the audience that might not be super familiar.

Joerg: 21:19

One of the core ideas of SORA is that we basically simplify the problem in boiling down each conops to fit into one of six dedicated operational levels. We call them the SAIL levels. They're the specific assurance and integrity level. And this has one major advantage because, eventually as an operator, you need to specialize on something, you know, there's a difference between a SAIL 2 and a SAIL 3 operator, the kind of amount of documentation they need on their organization. Um, if they need to be audited before they can operate, or if they just have to show what their operational manual looks like before. And so an operator could say, you know, uh, it is for my company as it is today, I can go up to SAIL three, I'm not going to address SAIL four because that would require me to restructure my organization, increase my overhead. I need to hire more people for certain safety roles that become mandatory at SAIL four. The same is possible if you are an OEM. You want to build a drone that is to be used for a certain type of operations. You can ask yourself, well, you know, I could make a SAIL 2 drone, which is fairly straightforward, or I could go SAIL 3 or SAIL 4, or even a SAIL 5, depending on what I market expectively. What do my customers want? Where do my customers want to fly my drone? The problem is, of course, uh, it's a trade off. The higher the SAIL level, the more expensive will the development of my drone be. So I have to hire at a higher price per piece. So that will affect the operation cost, of the aircraft by the operator. So, it's an overall trade off. The question must be, okay, for a particular type of operation, what's the right SAIL? Which one is cost optimal? And, I've already seen the first, um, operators that embrace this principle and, uh, start to think about, for instance, I have a drone that can fly over population densities up to, let's make, pick some number, 300 people per square kilometer. And, I can, deliver goods to so many homes. I would really like to go to these neighborhoods as well, but the population density is higher there. So the question for me as an OEM could be, well, if my customer wants to fly there, how do I get there? Do I, for instance, install a better mitigation system that slows the drone down in case of a crash? For instance, using an advanced parachute system, I could increase the effectiveness of that mitigation in my drone system, and by that I can fly into higher population densities by a factor of 10, and I can do it to more customers. Or, I could change the drone design, the way my autopilot works, the way the redundancies are being, organized in the, aircraft design and I achieve that and I get a more reliable, drone and I qualify my, uh, people better. I have a more robust organization. And both these approaches get me to something similar, with the downside that the parachute drone might actually, go down unplanned more often. The clear question here is, which one of these two scenarios is the more affordable one? And, or is the, the more economical? And this is not always obvious. Uh, it's a trade off, problem that, you know, first, drone delivery companies and OEM catering for these, companies actually started to think about. And I think, if there is a clear path to success in this, business of unmanned drones is to really look at the regulation that is in operation. Consider your constraints. Consider what you actually want to achieve. Think about there's multiple ways of getting to your goal, which one's the most cost effective way. Because, let's face it, if you have two competitors, and one can, do the same kind of mission at a lower cost, they will probably get a bigger market share.

Luka: 25:16

Can you give some examples in terms of use cases that correspond to the different SAIL levels?

Joerg: 25:22

I can try. This is not easy because the way we look at the size of drones, we don't have any weight limits in the system. So technically you could, go to unlimited size, which of course makes no sense, but there is no, nothing like a 600 kilo or, limit of that. So there's no, 25 kilo limit or 55 pound limit. We don't think that way. For instance, if you have a drone, let's say a two meter size drone, let's pick a weight, 15 kilos for instance, and you want to fly residential or sparsely populated delivery operations sticking to, unmanned cargo operations for now. You can most likely do that as a SAIL 2 operation. If you're not flying through any, densely used airspace, you're avoiding, airports on the ground. And, or if you have an airport in the vicinity, you at least have some sort of, organization, with the, airspace, owner. Similar to what LAANC offers in the United States, something like that. And, you could possibly fly these kind of operations as a SAIL 2. But you cannot fly into the suburbs with that. You can fly to, uh, for instance, single farms, in the countryside, but you cannot fly into, um, more congested areas with that. Now, sticking to that example, if you now make this drone bigger, what'll happen is that most likely becomes, most likely a SAIL 3 or SAIL 4 operation flying the same kind of trajectories. Also, what could happen is, you go back to the small 2 meter drone. Now we go into the suburban areas and what the system will do is say, okay, now you are going to, sparsely populated suburbs. Your SAIL increases to three. If you want to fly to the more denser suburbs, then what will happen, you will go to SAIL four, unless you can increase your level of mitigation. So for a lot of drone operations that we begin to see that are evolving in the, let's say, below three meter size range, below 55 pounds, roughly. What we will see that there's a sweet spot for, either SAIL 3 or SAIL 4 if you want to go urban, delivery. There's also the same, there's a sweet spot for, going with SAIL 2 drones if you stay outside of cities. And if you would go, take a big drone, like an eight meter drone and flying into the downtown area, most likely we'll get a SAIL 5 or SAIL 6.

Luka: 27:41

For SAIL 5 and SAIL 6, one needs a type certified drone, and when one has obtained a type certified drone, which I don't think that any OEM has to date, but how is that different from a certified category of drones?

Joerg: 28:00

The difference would be in that case, well, let's first of all, talk about, the JARUS SORA, because SORA is developed by JARUS, is, saying that the integrity of the design must be verified by the competent authority, aka the National Aviation Authority, that is responsible for you. And the easiest way of doing that is usually by using a TC. If you look at the EASA version of SORA, they clearly spell it out at SAIL 5 and 6. You need a TC, but it can be a TC with restrictions. So, the difference to, certified operations would be that in certified operations, you don't need to care about where you fly. You can use all airspaces, you can overfly any population density, you basically have to spend no thoughts on that. If you are in SAIL 5 and 6, there will be some operational limitation. For instance, you can fly over all population densities with this particular drone, but you cannot use all airspace classes, or you can make, might you be able to use most of the airspace, but you cannot fly over, population densities above 5, 000 per, square kilometer. That could be a restriction. So, you might need a type certificate because you have a very, big drone that poses quite a hazard, so you don't want that to crash. So a type certificate probably makes sense, but you don't need to bring it to a level that it can do everything. There's a good analogy, actually, if you, think about it that way. It's a little rough, but, bear with me. When, a SAIL 1 is similar to an open category or part 107 operation, A SAIL 2 is like a Part 107 or like a Part 107 plus waver, or, something like that, but with a very similar drone. A SAIL 4, I'm going to jump over SAIL 3 real quick. The best analogy is, that's a light sports aircraft, kind of, but as a drone that is being used for commercial operations. SAIL five would possibly be, a single engine certified manned aircraft like a Cessna 1 72. SAIL six is a twin engine certified is an airliner because, let's face it, this kind of risk, based approach we also have in manned aviation because the certification requirements for, Cessna 1 72 are not identical to an Airbus A320. So this logic is not entirely new. Now I skipped SAIL 3 let's go back to SAIL 3 real quick and then because this is really interesting. So if something, like SAIL 4, similar to light sports aircraft, that's a drone design that is airworthy, but the airworthiness is being declared by complying to, and everything is design standard. This, um, declaration is being made by the OEM. So there is no traditional certification process involved. The question is, what is less than that? Now you're allowed to, let's stick to the, SAIL logic. You can, at SAIL 3, crash 10 times more often. So which are the parts that you no longer verify? Which do you still need to look at? And SAIL 3 we didn't have this before. This is a new mode of operation. You have, let's put it this way, the aircraft that would be used in that category are airworthy in a sense, but not in the classical sense. And this is why it is often difficult to have the discussions and to initiate these kinds of operations, because this does not really exist in manned aviation. We don't have, drones that, at least on paper, or don't have any aircraft, at least on paper, are this unreliable.

Peter: 31:27

So Joerg, with respect to how companies are engaging with these SAIL levels and targeting which level they're going to go for. Are we seeing them start out at SAIL two or SAIL three and then over time walk their system up to higher SAIL levels? Or is that really more of a different product, a redesign effort?

Joerg: 31:53

From my point of view, if you want to go from SAIL 2 to SAIL 3, there's currently two options. The first option is, as you just put it, you redesign your aircraft to actually meet the design requirements at the most prominent here is you cannot have any system that is, likely to fail result in, a loss of control of the aircraft. Uh, so you need some redundancy. For instance, you cannot fly with a single battery. You need some redundancy in there so that if your battery fails, you can continue to fly.

Peter: 32:24

I'm interested if companies themselves have a reason to try to walk their aircraft up to higher SAIL levels over time, and if it's possible, and it really ties into the means of compliance, and if we're yet seeing convergence on what the means of compliance are for these different aspects of the integrity of the aircraft.

Joerg: 32:45

The alternative to, uh, redesigning your aircraft is technically to show operational experience that your aircraft is such a robust design that you can easily fly up to, 3, 000 hours without any major intervention. This is similar to the D&R process that was being used by the FAA in the past years for the first type certificates for drones. And these two very different approaches have both have their pros and cons. If you have a new development, it's probably more economical to, redesign a drone that actually meets the SAIL 3 requirements. If you already have a large fleet of SAIL 2 operations and you can easily create the required numbers, it might be more economical to go on a more test based method to show that your drone is reliable enough. But back to your question, um, do drone companies start out with SAIL 2 drones? In most cases, yes, they do. Now, this is where kind of, what I've seen with many companies is, or actually with most companies, they are building something that could be operated on a SAIL 2. The technical design requirements for SAIL 2 are very minimal, actually. And then they kind of are looking for investors and they're trying to sell their products. They look how great my drone is, how awesome looking, look at my features and look at my stats and so on. And then they're going like, yeah, eventually we're going to go for SAIL 3. Okay, sometimes I ask these companies, I just was speaking to a company, um, like half a year ago that I met at a drone event and they had this awesome vertical takeoff, a drone that could then transition to forward flight. It had really good, performance, data and it was a pleasant drone designed to look at. And, you know, it's something we're saying, well, you know, this can actually work. This is a nice design. So I asked them, okay, what's, what's next for you? You know, are we going to go for SAIL three? And the question that the answer was right away, yes, we're going to improve our drone for SAIL three. Uh, we know we're going to need it and so on because otherwise we cannot really get there. And I said, okay, how do you plan to achieve that? And the answer was, well, we don't know yet because we haven't really looked at the requirements yet. That I found really interesting. How is it even possible that you are setting up a new company, you know that you will eventually need to, uh, go there. And not really know what the requirements are, because yeah, yeah, we didn't find the time to look at it yet. And this kind of awkward conversations I've had many times over the last years. So whenever I, uh, talk to an OEM that says, Well, we have looked at it and for now, since we need to redesign our aircraft, we have chosen not to do it at the time, or we are already in the process of, setting up a new development, or we are going to jump sail three We want to go straight to sail four because we see our business case there. We are getting the expertise together in order to do that. That's the right way to go. Unfortunately, uh, I only see that in a minority of the companies, which is kind of due to the, let's, be frank here. The drone industry as it is today is still somewhat immature. We're still starting out. It's kind of weird because I've been doing this for more than 20 years. We've been this, we're about to take off mode for quite a while. But that being said, um, still, I believe we are in our infancy. These kind of drone companies that don't really know what they're doing will eventually disappear. I believe the consolidation is bound to happen at some point, but some of these new entrants will survive and will thrive in this new ecosystem.

Peter: 36:18

Well, certainly there's a long standing argument in engineering that safety and reliability have to be designed in from the very beginning and that it's very difficult to add that on later in the design process. And I think it's just interesting to consider that argument against the specifics of this drone application and where that holds true and where there are exceptions.

Luka: 36:44

So, Joerg, as we're talking about the different SAIL levels, and you're drawing parallels with the FAA, I think another parallel to draw would be that, the equivalent of the open category broadly would be part 107. The equivalent of the certified category would be, you know, this weird 21.17b path that the FAA, is forcing OEMs to go into in the absence of Part 108, But you know, a big gap in between. Whereas in Europe, this gap is obviously in the specific category and then further subdivided into six different SAIL levels. Do you think that the regulators have confused the industry a little bit with the granularity of the SAIL levels? To what extent has this contributed to, the maturation that we're seeing in the industry?

Joerg: 37:36

The specific category and the SAIL levels of SORA will be explored from the bottom up. You know, we are seeing SAIL 1 into operation at the beginning, then the first SAIL 3 started to come up, and I haven't yet seen a single SAIL 4 operation out there. And the same goes, of course, for SAIL 5 and 6. These are spots in the risk matrix that still need to be exploited. The biggest challenge for OEMs is, Uh, to begin, okay, they look at the requirements that they can derive from SORA, uh, for SAIL 3 and SAIL 4, and then they go to their respective authority, say, okay, but how do I show compliance to this? And for quite a while, there has been conflicting answers, and, uh, that led many OEMs to be a little bit more careful because, you know, you don't want to be the first adopter that spends the most money on this, and everyone else had it easier, but that's actually just part of the game. What I would expect is that the fog will eventually lift. We've already seen successful SAIL 3 operations in Europe and beyond Europe as well. Besides, you mentioned, your question was directed at how is it in Europe, but since SORA is being used outside of Europe as well, I'd say that is, equally applies to these other regions as well.

Luka: 38:53

Do you think that risk is sufficiently understood or is it conflated with hazard?

Joerg: 39:01

Well, I think you can find that quite often. Um, in a lot of discussions you will see about, you know. But people say there's a risk of that X happening. There's a risk the drone might crash into an airliner and everyone on that airliner dies. What people mostly talk about is to, in fact, the hazard becomes risk if you factor in the likelihood of that happening. So if this is a hazard that could happen, what you need to do in order to provide, uh, a sufficient level of safety is to make that event happening very unlikely. And, uh, summing up, the risk is actually low. So what we can do in, uh, the European rule system, we can fly drones of all kinds of sizes in principle. And it does not matter what is assurance integrity level your operation is at. If you have a very, complex, operation where you need, to pay a lot of, attention to, the level of airworthiness to the qualification of your crew and qualification of your operator or if you have, a very small drone that is flying BVLOS for a simple surveillance task, just flying for five minutes over unpopulated terrain. And bottom line is these operations, when they're being approved, have to be equally low at their risk level. Like, the risk must be acceptable to society. So the risk is always low.

Luka: 40:23

Let's, pause here because I think this is a really, really important, point to underline. You mentioned the same, operational safety target, so to speak, for any drone that is flying, in the air, right? Whether this is an open or specific or certified category, it's just a matter of how that safety target is reached, whether this is through design assurance, whether this is through, the operator certification or whether this is through the organization itself, or quite frankly, whether this is done via operational limitations, but ultimately the different SAIL levels really dictates the ratio of all of these different levers that ultimately get everybody to the same safety target. I think that this is an idea that is oftentimes missed in the market because what people refer to is in the SAIL level, for instance, the number of the SAIL level corresponding to the probability of a catastrophic failure. So for instance, for SAIL 4, it is 10 to the minus 4 in terms of probability of a catastrophic failure or of a given system function component or whatever, but that just as I'm interpreting this, this just refers to the design piece. But then there's all these other elements that get you ultimately to the same safety target. Untie this knot for us, please.

Joerg: 41:49

Okay, so your question refers to one of the core principles of the SORA methodology, which is the baseline that was developed by JARUS to basically try to get some structure into this big area of, different size drones, different operational scenarios. You have varying qualification levels of the crews and of the organizations. What this boils down to is if we know that the risk to third parties on the ground and in the air needs to be equally low, no matter what kind of operation we perform. It becomes quite, an interesting thought example because let's assume you have a drone that, would crash once every 10 hours. On average, you know, not only for design failures, but also because the crew is not really qualified very well, their procedures are kind of out of whack. So whatever they do, they crash every 10 hours. How do you have to limit this operation? What kind of operational limitations do you have to put in place such that the likelihood of killing a third party, somebody in the ground, is acceptably low? Can this be done? Yes. If you do this operation, for instance, in a controlled, ground area, restricted airspace, why not? There is no other aircraft flying around. There's nobody in the ground. Who are you going to hurt? When we, transition to, somewhere like to, let's say we go to a suburban environment, there is people on the ground. You can fall down and, you know, affect people. So the question is, given your drone size, what's the likelihood of hurting a person on the ground so that he, eventually dies of his, of his, contact with the drone? And depending on the size of the drone, it makes sense that the bigger the drone is, the more likely it becomes that you, hurt somebody on the ground. If you make the drone smaller, it becomes less likely. And if you know what kind of drone you fly over this environment, you can use statistical models on how drones or how aircraft in general usually crash. There are several good crash models out there that have been used in the past for range safety, for instance. And then you can basically say, well, you can fly over that suburban area if you don't crash more often than, I don't know, every 10, 000 hours or maybe every 1, 000 hours is acceptable, depending on the hazard posed by your drone. And then you automatically know if you need to have an equally low, level of safety as this example that you had over, you know, an unpopulated area, then you must be X times more reliable. So the operational reliability can no longer be, 10 hours, your operational reliabiity needs to be something around 1, 000 or 10, 000 hours, between loss of control events. And you get to the same, number for the overall, uh, level of safety, which has to be acceptably low.

Luka: 44:49

What is that universally acceptable safety target?

Joerg: 44:53

Well, what we use in the modeling, while developing SORA is we accept, one fatality every 10 to 6 hours, a single fatality, not a group of people, but that, uh, this is, for, a single person. When it comes to, hitting aircraft, it is 10 to the minus 7 for, uh, in uncontrolled airspace and 10 to the minus 9 in controlled airspace for, basically airliners which, we assume that if you hit an airliner, the airliner will crash, everyone dies. So if you have a similar order of magnitude for the safety target is 10 to the minus nine, and this is actually in line with what's being used by air traffic control today.

Luka: 45:30

And the important message here is, the 10 to the 6 hours for a single fatality on the ground, that needs to be achieved, by a SAIL 1 operation, as well as the SAIL 6 operation, ultimately, it's just a matter of how you. How you get to that

Joerg: 45:50

Exactly.

Luka: 45:50

With SAIL 6, it might be that you need to go through type certification to really approve the design robustness, because you're flying above a lot of people, you're less operationally constrained as opposed to a SAIL 1, where you don't need that robustness in the design, but you are constrained to very low population densities.

Joerg: 46:09

Yeah, that's the core concept. That's exactly like you just, summarized it. And, basically the internal values, we don't have to go into too much detail. For instance, the GRC number, the GRC number basically, uh, represents order of magnitude of the likelihood that after you lose control, somebody on the ground dies. So the higher that likelihood is, the higher is the GRC number. And this is on exponential scale, and the same goes for the SAILs. With every increase in SAIL we expect that the likelihood of losing control of your operation, that you crash somewhere, is reduced by a factor of 10.

Jim: 46:45

So, what's been the overall progress of the drone industry in recent years? Give us a perspective. We talked to Lorenzo Murzelli a couple of years ago, and he gave us a great perspective. What's changed in the last couple of years relating to the industry in Europe and globally?

Joerg: 47:00

I believe what's happening is that we're seeing more and more, operations that go, uh, towards large scale, either exception based in the United States or based on the SORA system in other parts of the world. What we're seeing is that the drone industry is starting to, mature. We'll see that certain use cases apparently can be done. Drone delivery is, uh, an obvious one because, you know, there wouldn't be able to scale up, drone delivery businesses if they wouldn't work. What we're also seeing is, the, a big spike in agricultural use of drones. And all in all, I think we are on a very good trajectory. What we're still learning how to use the ropes, kind of how to do, uh, how to navigate the new regulatory framework efficiently, but we're getting there. I'm quite, optimistic.

Jim: 47:50

What economic value will be unlocked with a, better understanding of or execution of the regulatory environment?

Joerg: 47:57

Well, I think the most imminent effect, if the authorities know how to apply the rules, uh, more efficiently, if, the drone industry understands the necessary means of compliance, better, what we will see is that, it is becomes less threatening for OEMs to move their designs forward to design SAIL 3, SAIL 4 and even SAIL 5 drones in the next couple of years. And, a lot more companies will decide to make the investment and go down that path. Because they will see successful operators that use these type of drones and they say, you know, actually I want a piece of that cake too. So, once again, once the authorities get more comfortable with the new system, we will, eventually the cost of operating these kind of drones will go down. And I'm not an economics expert here, but I would, forecast that we'll see some mind blowing developments in the drone industry. And we, you know, we can't really even say what exactly we're going to blow our mind, but, there are so many things that are possible either with, uh, using exceptions or, in those countries that have a commercial, BVLOS framework. it's an exciting, time to be alive and, you know, be part of that ride.

Jim: 49:14

So you've been doing this for 20 years and you're saying you can't really predict what the future is going to bring, but 10 years ago, let's say, what's happened in the last 10 years that you would have said would have been hard to predict 10 years ago, but happened and you're saying is a great development of the drone industry.

Joerg: 49:29

10 years ago, the idea of food delivery by drone was already being, discussed, but it was something to make fun of. Like, look what these idiots are going to do, right? I mean, that's never going to happen. I don't like, let me quote, you know, what we used to say, or what my colleagues used to say. I don't see that a package delivery company is going to deliver right to my backyard. you can make a great video of it, but you know, seriously, this is not going to happen at scale. Today, I'd say, well, actually, if you see these kind of operations, uh, first of all, the big change from 10 years ago, 10 years ago, this stuff was illegal. So, there was no way to obtain approval outside of a, you know, some exemption process where you are granted to do something that is otherwise, against the rules. Now we have a rule system that actually allows these things in many countries. So this is a fundamentally different situation. So 10 years ago, we'd say, well, again, I would have said, you know, we would have cargo drones. But they will probably fly airport to airport. Now we have a very different situation.

Luka: 50:33

You mentioned the agriculture use case. What's behind the growth that you're observing in this category and is this in spraying or is this in analytics where, geographically do you see the pickup?

Joerg: 50:48

I honestly do not know.

Luka: 50:50

Got it, but you know, that ag as a use case is, is growing. Okay. Well, let's have a hypothetical conversation. Let's say that we want to, deploy drones for spraying, for instance, and on the one hand, ground risk can be managed, fairly adequately if this is obviously in a rural area, but if it is close to, an airspace class that is, perhaps near an airport or in controlled airspace combined with the fact that a spraying drone, just has a payload of about 50 kilos. And so with the airframe, you're really starting to approach, a hundred kilo mark perhaps for the overall system, max takeoff weight. And so it seems like even though you're starting with an intuitively low risk operation, if you add up the air risk and ground risk, you end up in a, you know, sail three, possibly sail four situation quickly. What are some of the limiting factors that, influenced the growth of that use case?

Joerg: 51:47

The bottom line here is, uh, how, what to do with, other, manned, and even unmanned traffic. First of all, I believe that this time of agricultural operations, even with drones around the 100 kilo mark, if they're flying at low altitude, it's not so hazardous actually. There is of course, you know, if you are agricultural, you have crop dusters that are using the same piece of airspace, you either need coordination with them or you need to be conspicuous. Both, actually. So, what I believe is that, the conspicuity problem is, uh, one of the driving forces. I've seen developments of, or concepts for, uh, ADS, light, transponders that, you know, are operating at, lower, power, that could, help, bring conspicuity. But, in the end It must be affordable and, otherwise you will have an issue with scaling. If you have a one off operation anywhere, uh, usually you can kind coordinate, uh, the airspace use. If you, want to scale up, this is most likely not, uh, possible. And, to go to the next, difficulty factor here, if you then go to, areas where a lot of people want to fly, if, as I, you know, my vision of the future, everyone will use drone delivery. That means suddenly we have a lot of drones over every city. And also we have airports in those cities because people like to, uh, not drive for an hour and a half to the next airport. They want to have an airport at their city. So, bottom line is that, in these kind of situations, this is where, uh, the developer of UTM comes in handy. That's, the solution to that, scaling problem because, you know, you can fly, uh, an, a drone, in, an urban area. You can fly close to an airport. All these things have been approved in the past, but scaling these operations up to having hundreds of thousands of drones per city, that will, uh, take additional combination. But it's not necessarily that the SAIL will go up for these kinds of operations.

Luka: 53:45

We opened the podcast with a conversation around air risk and detect and avoid and airspace integration, mitigating air risk is one of the big emphasis for the upcoming new iteration of SORA 3.0 can you speak to that a little bit, give us a glimpse of how, the gray beards at JARUS are thinking about solving this problem.

Joerg: 54:10

So JARUS has released Sora 2. 0, which was the first really usable version, uh, in 2019. It got adopted by, several countries across the globe. The most prominent was, the family of the EASA countries, uh, but also Australia, New Zealand, Israel. I've heard, that it's being used in Georgia. There are modified versions of Sora being used by Indonesia, Ghana, and, Canada. Um, so, it has seen widespread adoption. And when it was released in 2019, it means it was developed in the years 2017 and 2018. And with respect to the air risk model in that version 2. 0, that was developed there, it was a very conservative model. It's better to be safe than sorry. After all, it was developed by aviation authorities, and they are usually quite risk averse. But 2017, the discussions we had in 2017 for the drone industry today, that's the age of the dinosaur. Um, back then we had the idea that UTM would be available in the future. But there was no such thing as a UTM implementation anywhere. There was no rules for it. There was nothing. So, when this was written, it was a very conservative, assessment of the air risk and it was, let's put it this way, it is not really easy to use when you're new to the topic. When we, uh, released this, there were still some pieces of Sora missing, so it became obvious to us we wanted to do an update of Sora. And especially since it has been, applied, by many countries, we, uh, wanted to learn from the application. So the missing pieces of SORA should go together with an overall update based on the lessons learned from our new application. But since this can become quite complicated and, our authority members don't have endless resources to do this kind of work, it became obvious after like a year of work that we would have to split this update process into two halves. On the way from SORA 2.0, we would do an intermediate version, which was just released this year as SORA 2. 5. Now, SORA 3. 0 is literally the second half of this big update. And what has not been updated in SORA 2. 5 is the Air Risk Assessment portion of SORA. So the big focus will be on getting a more realistic and more easy to use Air Risk Assessment system into the SORA.

Luka: 56:37

What might that look like?

Joerg: 56:38

I don't want to spill the beans here because in our recent meetings the first concepts were being developed and discussed. So I'll hold back and, we'll not talk too much about it, but let's put, let's talk about the requirements for the update. What it became obvious to us is that the airspaces we designed, uh, it for, or what the, what the kind of, what we were anticipating, uh, is based on the airspace realities in Australia, New Zealand, Europe, and, North America. Many more countries across the globe and their airspaces have their own unique features. Some have, uh, very low traffic numbers or some have, very little, general aviation, but, uh, massive airline operations, some have a lot of offshore operations, lots of helicopters. And they're not all the same. And the Sora Air Risk model in version 2.0 and 2.5 is a, is a try to one size fits all. And that's also the downside of it. I'm not saying that it's wrong. It's not wrong, but it's often cumbersome to use. Let me give an example. If you want to operate over an urban area, the Sora Air Risk model assumes that you have, uh, what we call an Air Risk class of C, which means that you should at minimum be a SAIL 4 operation, unless you can mitigate that Air Risk. This was been put in place because there are some regions on the world where in the congested areas of cities, you would expect to have a lot more low flying traffic. There might be police helicopters, there might be EMS operations, and that was the logic behind that. Now it's being applied, for instance, in my home region in Germany, where we have a lot of smaller cities. But if you use the SORA Air Risk Assessment System, it says, well, is it an urban area? Yes. Okay. So we expect that there's a lot of helicopter traffic around. While the reality in most of the cities that are usually of smaller size, but we have more cities, is that, it's, it's not really that congested. It actually is more busy outside the city because that's where the sailplanes, uh, the gliders do their emergency landings or their off site landings. So there's a lot, there's actually more traffic outside these cities. So this one size fits all from my home region, it does not work that well. Of course, SORA then allows you to do individual mitigation. You can argue that over this particular city I'm flying over, there's not that much traffic. Problem is, how do you prove that? If you go to the ANSP, they will say, Well, we don't really have any radar coverage that low, so we cannot tell you how much traffic is over that city. Well, I mean, you could argue that, well, obviously, if there's no radar coverage, there's no need to have radar coverage down there, because there is really no traffic, or very little traffic, at least not enough traffic that it warrants to have radar surveillance. It's kind of like a, almost like a chicken and egg problem. And while this often makes sense, it effectively limits the usefulness of the SORA in its today state because often you have to enter discussions with your authority and your ANSP again and again. And one of the ideas of the updated air risk model is that, you know, first of all, there's more options in that model, in the flowchart to determine your actual air risk class. We're going to integrate the most popular mitigations right into the flowchart. Uh, so you don't have to, think about it forever and, uh, have to understand, let's put it this way, sometimes it's often cumbersome to use concept to mitigate your air risk into something more straightforward. Actually, for instance, in Europe, there is some guidance material being published by EASA that say, how do you usually mitigate your air risk in certain typical situations like the one I just described to you. And based on that, a lot of the authorities in Europe are basing their decisions. So with this additional knowledge, how what the best practice actually looks like, this should influence the next iteration of the Sora air risk model, so that you will have something that is not as alien as this current system that again has its roots back in the year 2017.

Luka: 1:00:47

Back to your example, you're flying over an urban area, you're automatically in Air Risk C class, and if you don't have the benefit of UTM, which would put the risk into the B class, and you don't do an assessment of the airspace to show the regulator that, hey, really the chances of an air encounter is low because aircraft are really not flying here. And, you know, here's the result of our, month long analysis of the airspace. So if none of that is done, and in fact, you are in ARC C air risk. Are there even sensors, methods, techniques available that will allow somebody to operate in that airspace. Because so far we don't have a SAIL 4 approved systems. So far, we don't have a medium robustness, tactical mitigation, for air risk, can you shine some light as to how that will be addressed? Will you be using what is available today in the industry or is it up to the industry to develop new technologies?

Joerg: 1:01:53

It's definitely up to the industry to develop these technologies and to the industry consensus standards or the standardization organizations to develop stuff that actually works to deal with this problem. But in this particular example, you're talking about, well, let's just accept the fact that you're a SAIL 4, you need to have the medium robustness, tactical mitigation solution that the SORA prescribes. And, is there developments on the market that could allow that? Yes, I believe these things will happen in the next couple of years. It will become more and more common. But then, thinking about our little toy example here, if this is really a small town, what's the point? Why do you need to be SAIL 4 and have this amazing detect and avoid capability if you're flying at an altitude where you basically, it is very unlikely to encounter any other air traffic if you stay away from helipads. And those are usually known where they are. The more interesting problem. So what in fact happens in a lot of these situations, in bigger cities and more congested areas where they are fielding UTM systems because they will eventually need them, it is, becomes easier to get an approval for drone operations below 500 feet than for small towns that actually nobody cares that much about because there's less business to be made.

Luka: 1:03:06

Do you, okay, yeah, that's a good point. So do you expect that the Sora 3. 0 will rely heavily on UTM or the availability of uSpace? Or can we expect to receive some very specific guidance in terms of that you need to be able to account for, you know, 80, 70, 90, whatever the percent of traffic is within a certain field of view, at a certain distance from the ownship.

Joerg: 1:03:32

It will be, exactly like that, we will, um, first of all, what will most likely happen is that, uh, it becomes, as I said, easier to, digest. It will have a more realistic initial, risk assessment step that ends up that, for instance, small cities most likely will not be classified in the wrong air risk class. So that already takes a lot of, uh, pressure, uh, right away. Speaking about UTM, integration of UTM will become a major part because UTM is here to stay. It's not going to go away. But let me highlight one little thing about UTM. UTM stands for, uh, Unmanned or Uncrewed Traffic Management. So what's the prerequisite for actually using UTM? You need traffic. Because it manages traffic. When you don't have any traffic over a particular area, you don't necessarily need a UTM system. In fact, if you have low traffic numbers, you never need a UTM system, unless an area is so interested to be operated in, or so close to a major airport, that there is a lot of traffic, because so many operators want to use it, and that is where you definitely need it. Think of UTM like a traffic light at an intersection. Are traffic lights bad? No, they're fantastic. They're a great solution to a problem. A problem of too much traffic, too many people want to go from place A to B and they have conflicting trajectories and the traffic light sorts it out. Do you need a traffic light on every intersection of two country roads? Not really. There's other ways of mitigating that risk. For instance, you could, you know, carefully approach that intersection, look out, and if there's nobody around, continue your way. What I sometimes, I get the impression in some of the debates over UTM or uSpace in Europe, not generally speaking, but sometimes I get the impression is, Like, we're going to install this, you know, new U-Space or UTM system in that remote area, and when we install this, traffic will come. That's, the analogy would be, we have a country road intersection, we're going to build a traffic light, and then we know that more cars will come. That's not going to happen.

Jim: 1:05:36

Isn't there a little bit of a chicken or the egg though? That the volume may not come unless a UTM system is deployed?

Joerg: 1:05:42

Honestly, I don't believe that. Because if you have a city, and you have very few operations in that city, the likelihood of these operations crashing into another is very low. So why would you need a UTM system to fly there? If you use part of a CTR of an airport of a, Airspace Delta or Airspace Charlie of an airport, it depends. There's the possibility of negotiating with the ANSP for instance, there's a particular section below 500 foot where you can get a general clearance, or you can use a system like LAANC or something along those lines. If there are very few operators, this is still a viable option. If it becomes many operators and they have conflicting trajectories, that is where UTM starts to shine. So will we need UTM? Yes. But do you need UTM first so that we can enable BVLOS operations? I believe no, we don't.

Luka: 1:06:33

So there might be a, sliding SAIL level or air mitigation depending on how much traffic there will be in a given area. How does that sit with an OEM or an operator to know that, Hey, you receive an approval, but as soon as this, market starts attracting new traffic, you might have to go back and get additional approvals or the ones that you have are not necessary anymore.

Joerg: 1:06:56

So let's assume for a moment we have the updated air risk model, we start out, uh, urban air deliveries, we are initially an air risk class Bravo, which Basically says that you can be SAIL 2 or, uh, or higher. And, so most likely this is going to be a SAIL 3 operation. As I said before, that's probably the sweet spot, the most effective, way of, flying in this area. And what happens is more and more drone traffic comes, and, you know, it should be coordinated at some point in time. And what you could do, what would happen then is, well, if you need to introduce a UTM service in that area because of the increasing traffic numbers. The con ops of this drone operation needs to have an interface to UTM. So the update that you need to do is, you need to, you know, either have a UTM because the air risk class is going to increase slightly. So either you have UTM, or you have a better outfitted drone as a possible solution to the problem. And most likely, I would say the most effective way would be to basically connect to the UTM service. In most cases, it will be mandatory anyway for that airspace from the system designs that I've seen so far. And I don't believe that your SAIL is going to change. You might have additional tasks that need to be put into your operational manual, uh, to, but I don't think that the SAIL will increase.

Luka: 1:08:19

If in the future, there will be more prescriptive performance requirements for air risk mitigation to meet, for instance, you need to be able to detect 80 percent of the traffic, two miles out. I'm just, hypothetically speaking, does that necessarily mean that you have to understand what those performance levels are for an average pilot flying a general aviation aircraft before you can start applying certain standards on the drone operators. And if that's the case, is that a known statistics?

Joerg: 1:08:54

That's a difficult question because, uh, whenever you want to talk about, these kinds of statistics is, okay, how good is, the manned pilot, that is, you know, obliged to, see and avoid, uh, other traffic. In my piloting days, I, um, realized that, you know, actually there's a lot of traffic that you don't see. But still, you don't have any accidents, or not many accidents, thankfully. So, it is really difficult to obtain real data, what the actual collision risk actually is, to begin with, and what the percentage of traffic you need to see. So, when we don't have actual data, the only choice we have is to revert to models. That's what we do in SORA 2.0 and 2.5 and we will have an updated model for SORA 3. And, with the new developments, especially on the, standard development side, there's also some, work at ICAO on, uh, on this topic on the required risk ratios for, detect and avoid performance. What it boils down to that we will probably have to adjust our, percentage numbers slightly, but not by order of magnitudes. And it will be, very well aligned with, what's developed, at the standardization organizations. I believe there will be some, detect and avoid provision as part of the upcoming part 108. That will also rely on industry consensus standards. So I believe we have not converged yet, but I see converges in the next two, three years.

Luka: 1:10:14

Okay. So timeline for Sora 3 is two to three years.

Joerg: 1:10:19

Definitely.

Luka: 1:10:19

Okay. So that's a long time. How do you think part 108 will approach this detect and the avoid issue?

Joerg: 1:10:25

I'm not going to speculate much about that. Um, let's, let's face it. Part 108 is something that we don't really know anything about. I barely know people that have actually read the drafts and the ones that did read the drafts because they are members of the FAA staff that has to be developing it and they're not talking about it. So right now we don't really know much about Part 108 we do know something about the characteristics of Part 108 it will definitely have a solution of the technical mitigation problem for the airspace. Definitely. What we also know is that, uh, Part 108 will, put a lot more emphasis on the responsibilities, and liabilities of the OEM and the operator, which means you can declare compliance, uh, to certain, to a lot of the requirements, or you can declare compliance to the awareness requirements and so on. And these things are actually very, have a lot in common with, the SORA framework, if you think about it. Part 108 to me sounds a lot like what we do for SAIL 3 and 4. Will it be completely identical? I doubt that, but in the orders of magnitude game, that is where we probably will find, Part 108. Which, uh, provides us with an interesting opportunity. Can the requirements that you have to comply with for, for instance, the airworthiness of the drone and for Part 108, and the ones that you have on the SORA side for SAIL 3 and 4, Are they compatible enough that we could agree on the same means of compliance to achieve both? I believe that's an opportunity to actually build a bridge, to the system. Because if you look at the engineering logic in part 108 and what we have in the specific category countries, the underlying logic is actually quite similar. If, if you're looking at, FAA, what's it called, order action, 44807. Some of the things that are being mentioned there, that's the basis for the exceptions and the current FAA system, pre 108 That's logic in there, it's actually quite similar to what we are using in the SORA countries. Not identical, but you know, at least we're not living on different planets, literally.

Peter: 1:12:34

What is it about the FAA, framework that might cause part 1 0 8 to be, in some ways very different from what EASA has done with the SORA?

Joerg: 1:12:46

So, JARUS is the Joint Authorities for Rulemaking of Unmanned Systems. It's an organization that is now 16 years old and it's basically a voluntary organization of, I think we are around 70 aviation authority members that are basically meeting multiple times a year to, have discussions on, how to develop, rules for the use, of, unmanned aircraft systems. And, the most important product of JARUS has been the idea of creating, what we now today know as the specific, uh, operations category for drones. And, our most, famous product, is not the most important part, but the most famous product is the, SORA methodology, the Specific Operational Risk Assessment that has been developed by, the work group that I'm now, privileged to lead, the work group for safety risk management of JARUS. And in JARUS, we have a mix of, uh, roughly two thirds of the members of the workgroup that develops SORA are, from aviation authorities, aviation authorities officials, or people nominated by the aviation authority to act on their behalf, and, roughly one third is, subject matter expert from industry, that are supporting the activities there, but they have no voting, rights. So whenever we speak about, uh, what the SORA allows and how what the SORA does, this is the, joint development or the joint effort of, both European, American, Asian, and, African. Basically the entire world is somehow represented in that, undertaking. That's why, you know, what EASA has done this, EASA has done that. EASA is adopting Sora. So is Australia. So is New Zealand. And to be frank, my good colleagues from New Zealand, especially in the early days of SORA have contributed massively to the success story it has become. So, you know, something to shrug, because EASA has, first of all, they invest in SORA. They helped develop it. Uh, the European member nations of EASA helped to develop the SORA and EASA was brave enough to actually adopt this as their means of compliance. That is what, uh, EASA has done and what they have done. Thankfully, done. But they're not alone. And I often find it difficult, like, it's like, this is the EASA system. Actually, no, it's not. There's other countries, for instance, the semantic model that we use in the SORA today was quite influenced heavily by a very smart, a former pilot, who was a, member of the South African Aviation Authority. He helped to get the SORA semantic model to what it is today. Just to give you one example.

Peter: 1:15:21

So what is it about the FAA's existing frameworks and legacy that would logically lead them to create Part 1 0 8 in a way that it significantly diverges from the SORA and from the direction that, you know, Europe and all of these other countries in the world are going.

Joerg: 1:15:40

Well, does it though? We don't really know that yet. First of all, I assume that the part 108 will bear quite some resemblance when it comes to the core principles. the US was, involved, the FAA was involved in developing Sora in the early days, uh, but, a couple of years ago, they have decided to, follow a different path, which of course, uh, you know, every country, every JARUS member can do whatever they like. You can adopt a JARUS product in full. For instance, Australia is a great example. They use the JARUS SORA exactly as is. There are some countries that have adopted the SORA to, you know, to suit their needs. There are some that have heavily modified SORA, like, Canada is a good example, or Ghana have, tailored it quite a bit. And there's countries that, influence their decision making based on the underlying logic. And, SORA was also, you know, I mentioned LSA to be somewhat similar to SAIL 4 operations. SORA was also heavily influenced by the FAA and their, the process that led to the creation of light sports aircraft, for instance, with all the discussion about the Mosaic program and, these things. There has been considerable influence, in both ways. I would expect, the development on the FAA side to be, in principle, somewhat similar. It will probably not be identical, but, I believe, if the core concepts are similar, we can, uh, find a way to bridge those gaps between the systems, because let's face it, do we really want a world where basically there's four different ecosystems for drones, and if you're an OEM, you can decide which one to develop for? Or if you're an operator, I specialize in, you know, uh, just in the Asian market, but I don't go to Europe because they're weird over there. That doesn't make sense. Nobody wants that.

Luka: 1:17:27

Totally. I mean, that's one of the beauties of Sora in being a common language more than anything. Anything else that we haven't asked that we should have? Anything that you're burning to share some beans that you're burning to spill?

Joerg: 1:17:40

There was a question about who adopted SORA 2. 5. I could answer that. No one has yet. We just released it in June, and I haven't heard of any successful applications by any authority. On the EASA side, they are planning to make it part of their official means of compliance in the first quarter of next year.

Luka: 1:18:00

And when we look at Sora 2. 0, there was, what, maybe 11 months between when Jarus completed it and when EASA adopted it. Is that roughly the timeline that you expect for 2. 5 as well, or?

Joerg: 1:18:14

Yes, roughly. I know that the draft versions of the EASA 2. 5 already exist and are currently being debated. Because, Europe, chooses to tailor it to their particular, framework. Europe is kind of unique, because we have, one agency that is the responsible authority for all airworthiness, topics and certification of aircraft, while the, responsible authorities for, licensing and for organizational approvals are the aviation authorities of the EASA member states. So, because of this, split European version of SORA has to reflect that. It has to say which authority actually gets to decide and to approve what. And in another country like Australia, they wouldn't have that problem. It's the same, authority. It's the same, person most likely that's approving all these things.

Peter: 1:19:00

If we compare from 2020 until today, we have not really seen that the path for a new drone company to go from clean sheet, through design, through getting approval under Sora to really have accelerated yet. And in part, that's, I think, due to a still maturing supply chain. But from your perspective, looking ahead, do you see gates opening up? Do you see convergence in the means of compliance that are going to enable new companies a year from now, two, three years from now to have a much faster path to get into the air with their platform? Um, will that fast follower opportunity for, you know, for drone delivery companies or for something else emerge, or is it still going to be the long road that it is today?

Joerg: 1:19:52

Now I believe things can be done a lot quicker in a couple of years, because now we know what the most likely means of compliance are. There is more and more guidance material that is being published by the countries that use SORA. And I think the bottleneck is more on the industry side. still, we get, still get to see a lot of, um, new entrants in the industry that have not really done their research, that are starting with great drone designs, they have great mock ups or even flying versions of their drone, and they haven't even looked at the rules, that, that they would need to follow, what kind of design, thing or which requirements would influence their design decisions. If you had asked me five years ago when we introduced SORA in Europe and other places, I would have said, industry is going to take off immediately. This is such a, you know, a barn door being opened, a barn door, you know, and the only opportunities lie behind. And interestingly, the majority of companies didn't go through the door. They just stayed on where they already were. They already had their customers. And that was, for me, the biggest surprise that I could have not predicted, that this kind of, behavior by the, new entrants of industry is actually possible. To be fair, those that went through the barn door then often met with, all members of the authorities that are still also learning how to use Sora for the first time that have a lot of different interpretations. And, one of the biggest problems that we've encountered over the last years we learned is that we wrote Sora with the best, experts that were available to us at the time. But we could not make up our minds how many misinterpretations and different, interpretations of the, provisions of SARA could actually be created. And that, of course, is a big problem. SORA can do great things in principle. Yes, it can approve BVLOS operations. It can prove, drone operations above, 25 kilos. It can do, fantastic things, but the market's never going to take off, really, before we actually all agree on the means of compliance and they actually apply them the same way. And this is still the, what we are, the phases we are currently, uh, The good news is, compared to five years ago, where there was basically almost no means of compliance, if you look at the older version of SORA, you can see a line in the annexes B and E, for instance, that often say, has to be developed to a standard acceptable to the authority. That is a placeholder that we put in there, because when we wrote it, there was no standards. There was no nothing. Now, what happened in the meantime is that the standard development organizations have created a lot of suitable standards that could be used within the SORA process, but the SORA doesn't refer to them, because they didn't exist back then. Now, some authorities, EASA certainly is now accepting certain standards for certain provisions of the SORA. But we're still, quite in the early days we have created a new, task force, within JARUS to look at this problem in particular that is going to take care of updating SORA 2. 5 with new developments. Because, you know, SORA 3 is going to take us at least two to three years before, it, will, become usable, but in the meantime, there are so many good new developments that are happening that we believe that we should have a discussion with the JARUS member, authorities and, discuss, okay, can we do intermediate update of the requirements section of SORA 2. 5. Like, have a 2. 6 or 2. 7 version that gives out more guidance material, has clearer references to available means of compliance and so on. And that alone will simplify the process because imagine you're going to, this is year 2020, you want to do a SAIL 3 or SAIL 4 operation. You go to your authority and say, okay, what is an acceptable airworthiness design standard that I could use for SAIL 4, for instance? And all you get is crickets. Now, what do you do? EASA has done their homework. They've created the the special condition light UAS, which is exactly going to serve that purpose. But if you look at the special condition, again, it has a lot of sub requirements. And for the sub requirements, again, you need means of compliance. Now you have to enter a discussion with, uh, the authority. Okay. What could we possibly do to meet this requirement? Again, the standard development organizations are, you know, tailoring new standards for each of these requirements. And you, of course, use a different, ever increasing design standard. I'm quite sure ASTM for instance, has developed something in the past. And they will definitely develop new standards, how to design, properly design a drone. These things are happening. This will make it easier. So, um, the barn door was open, but there was no road on the other side. Let's put it this way. So, you can walk over, but it is more burdensome. But now that these things are happening, that we are having the necessary debates on means of compliance, that we have to do, the emerging of new standards. I believe it's become easier and easier year by year.

Luka: 1:24:56

Excellent. Well, thanks, Joerg. Really, really fantastic conversation, and thanks for going into the details. Anything that you would leave the audience with as final thoughts?

Joerg: 1:25:07

I've already talked about this, the part where I, I mentioned that, it's going to affect our lives. The way we live our lives will change. It has potential to do that and, I'm looking forward to, experience that firsthand. I think this is actually fantastic, you know, what, what's happening and, this doesn't happen too often,

Luka: 1:25:25

Totally agree. Great time to be in the industry.

Joerg: 1:25:28

uh, most definitely.

Luka: 1:25:29

Great. Excellent. Well, thanks again, Joerg. Appreciate you coming and speaking with us. And, looking forward to doing it again sometime soon