The Ambulance of the future will fly, but will it use hydrogen?


I'm fairly convinced that soon we'll see ambulances switch to e-VTOL flying machines. So many advantages, hard to see downsides. Nobody is going to complain about noise and privacy issues of an ambulance. This announcement by an e-VTOL company and an air-ambulance company of a collaborative project is thus interesting, if preliminary. However, it's also interesting that they view hydrogen as the fuel. H2 has lost in cars, but has some positive attributes for planes, particularly an ambulance.

Read today's Forbes site column on this project at The Ambulance of the future will fly, but will it use hydrogen?


Wouldn't pure battery electric propulsion be fine for such a short range ambulance? There you mostly need high specific power, not that high specific energy, and batteries are somewhat better then fuel cells in that area. But even if you go hybrid, some kind of ICE solution seems more suitable. Either a high performing piston engine with a generator, or perhaps a free piston linear generator design. You'd get much easier fuel handling and the drawbacks of noise, emissions and sustainability, due to either using fossil fuels or biofuels with limited scalability, wouldn't really matter in such a niche application.

In any case, while ambulances seem like a no-brainer for eVTOL technology, I've also been wondering if the easiest first market might be for a vehicle that quickly delivers a paramedic to the scene with some basic equipment. If you primarily recruit among shorter women you might get away with a payload of only around 60 kg or so, which could enable a really compact vehicle that could land almost anywhere.

I don't really know how important getting a paramedic to the scene is in comparison to getting the patient to the hospital though, so it might not make that much of a difference.

I personally think you're right, that batteries would offer many advantages. However, one thing they would not be able to do is go all day. The hydrogen vehicle could operate pretty nonstop, while the battery one, after 30 minutes of operation -- possibly just one return flight -- might need 40 minutes of downtime and a very fast charger.

I have not seen comment on why they selected hydrogen for this. Usually the advantage of hydrogen or other liquid fuel hybrid designs is range. In fact, it is the only way to get range these days. An ambulance doesn't need range (unless it's a rural one) but it is better if it has less downtime.

To me, battery swap seems like a viable solution to that problem, but I've not seen a single player in the eVTOL space using it. Do you happen to know why?

My guess would be that swap hardware adds weight. Swap is OK for fleet operations, though not for private vehicles.

Have you even designed an A/C? Every last ounce of weight is painstakingly analyzed to benefit. No wait is added that doesn't bring benefit to the design. The specific energy of the powertrain is the most important of all, as the power train is what provided the energy for lift (so every ounce of powertrain weight is more weight that the powertrain must lift).

Hydrogen has a specific energy 3x higher than Jet A, so it is in many ways the perfect fuel for A/C. The only way it isn't perfect is in the volumetric density which is 3x lower than Jet A, but that has a lot to do with trying to jam hydrogen into a form factor that was designed for Jet A. Wing bodies are suited to hydrogen

Everybody looking at hydrogen is very aware of its high specific energy, that's why they are interested in it. But there is no way to store and use hydrogen that doesn't weigh many times the weight of the hydrogen, removing that advantage. If somebody can solve that ... fantastic.

It's not like many of your customers have the ability to shop around.

VTOL ambulances are already in operation, they are called helicopters. There are reasons for their low numbers, and for their operations primarily on remote calls.
1. There is not enough helicopter pilots, and never will be. The cost of training and keeping one is uneconomical even for such niche case as an ambulance. Few pilots means few helicopters, few helicopters means manual assembly and high unit prices. There is no reason for a helicopter to cost much more than a ground vehicle of comparable mass. The root cause of the order of magnitude difference in cost is pilots. Ergo, remove pilots, remove one zero from unit cost (eventually). Autonomy, no-nonsense hype-free bet-your-two-kedneys-reliable autonomy is the key for enabling flying ambulances and all kinds of other flying vehicles, which is what Airbus is dreaming about.
2. Noise. One ambulance helicopter flying low over a city at 3am wakes up at least a million people. Two of them is a good reason for an uprising in the morning. Hence the noise level of any VTOL must not exceed status quo: a truck or a bus, which is between 60dB (loud) and 80dB (too loud) at peak. That is also why helicopters over city are so few: large rotors generate low-frequency noise and infrasound that cannot be effectively screened by anything except distance. That makes any commercially successful VTOL small by necessity, as anything large will inevitably be too noisy for a city, and numerous VTOLs will simply not be tolerated.

In summary, based on facts: proven autonomy before any chance of commercial success, then small VTOLs with any power source. None of this is new, and all of this is already fairly well understood in this nascent industry.

Surely the job of a helicopter pilot will get easier and easier over time. Already the main expense of training for a private pilot license is the cost of renting and operating the helicopter (or airplane) in order to get the requisite flying hours.

Already we have airplanes and helicopters that can pretty much fly themselves in most conditions. The pilot is there because the FAA won't approve the flight otherwise (at least not without very expensive redundant safety features that can handle a catastrophic system failure).

One of the hardest parts of flying is dealing with ATC, and that system is utterly antiquated.

Helicopters are complex to fly, big, very noisy and can only land in very limited places. e-VTOLS are planned to fly themselves or be as easy to fly as a personal drone (easier, actually.) At first though, a human to guide them into tight spaces is nice. The computer will indeed handle everything but takeoff and landing at the patient location, and conversations with ATC.

e-VTOLs are much quieter than helicopters, are are down to leaf blower levels, which is fine for an ambulance.

There is no room for guessing what will be and how, now that there are ~20 years of statistics for drone operations and crashes. All those vehicles are inherently more stable aerodynamically than any quadrotor, and designed to a much higher reliability standards than any commercial eVTOL. The data is open:
In short, 64% of crashes happened mid-flight, 8% on landing, 7% on take-off. That covers 'piloting' any drone designed within the prevailing paradigm of assisted piloting. If link is removed and pilots are internal, which ruins the business case for reasons I described before, it is back to the current pilot-autopilot combination, which failed to produce mass VTOL transport since the invention of helicopters.

On the matter of noise, consider the case of Lilium - the closest eVTOL to reality and commercial use, also the closest to the ambulance case. In short, they are not ready to discuss that matter.
"Notably absent from Lilium’s video release was any accompanying aircraft sound. That’s significant, because most industry observers agree that low noise levels will be critical for public acceptance of eVTOL aircraft. Walker-Jones said the absence of sound was not a deliberate attempt to conceal the aircraft’s noise profile, but instead related to the difficulty of compiling a truly representative soundtrack from multiple distributed cameras. The company plans to share representative noise samples in the near future, he said."

Another one is EHang, the minimalistic quadrotor design, much smaller than Lilium. They are less shy on the subject, and their noise numbers are bad.
"The EHang 216 is equipped with eight sets of rotors which emit a noise level of 90 decibels, below a normal helicopter but still easily loud enough to be uncomfortable. The firm says it hopes to get the noise level down to 75 decibels."
Even 75dB is far too loud for any city, at any time of day. For reference, I measured 80dB standing 30ft away from a working mid-sized excavator with jack hammer. One tiny EHang is currently twice as noisy.

Uber's required noise limit for their flying taxi is 62dB at 500ft, or even 50dB at 40m (various sources). Surely those guys did a very comprehensive research, including public and legal, on what they could get away with.

On the subject of leaf blower people, the only reason they can do it more than once is because other people generally prefer not to be incarcerated. This is the only reason. Uber case study is a far better guide for what is acceptable in a city.

Correction: 20% crashes on landing, 8% on take-off.

There are none. And there are very few who believe Lilium is the closest to reality (most people I talk to think it's the furthest!) EHang, which also has ambulance plans, is closer to commercialization.

Noise efforts are in two areas -- takeoff/landing and regular flight. Most designs (not eHang or the one in the article) are fixed wing a short time (15 seconds or less) after takeoff. There, they are much quieter and vastly quieter than a helicopter. I believe that only hybrid fixed wing/rotor designs are viable for general use, but an ambulance has more leeway.

Yes, Uber has done a lot of research, they have people devoted just to that, I've see two talks by their main sound researcher. There are still challenges.

On the other hand, somebody I respect greatly says he has his vehicle so quiet that people inside buildings, with the window open, do not notice it landing next to the building.

Wings require speed to be useful, and speed implies significant flight time. City ambulance will have several minutes to fly across half an average city, and large cities have multiple hospitals, which keeps the flight short. There is no time to reach and hold significant speed for wings to be worth the trouble. Certainly all this would look more convincing with proper calculations, but I will leave that for another time.

Regular flight is of little concern, although it seems the solution for integration of all that new traffic into existing air traffic control rules will be squeezing it into the unused space between some minimal altitude and the floor for commercial aviation. That is roughly between 150 and 450ft, at or below the tops of tall buildings. That means low speed - no benefit from wings, forcing any VTOL essentially to almost hover. If they are noisier than 60dB, they will be banned from built areas very quickly. At 60 they are as noisy as trucks, which is a defensible case for city. That will not be tolerated in sleepy suburbs, with legal or illegal 'anti-access, area denial' measures put in place by the locals.

There are ways to make VTOL quiet enough, but it has to be small. That is not an obstacle for the taxi case, as the optimal load is one passenger. Ambulance requires at least two paramedics, at least one patient, and 400~500lbs of equipment by the looks of it. That means at least 1200lbs load, about 3~5x take-off weight, meaning about 500~800 kilowatts into motors for taking off, and HFC generating about 800~1300kW of power (about 1000~1800hp), 40% of which will be heat. All that power will be going into making wind at ground level or low altitude. If a VTOL is not small, it will not be allowed in cities, and ambulance cannot be small enough. A VTOL patient transport can be small enough, meaning a VTOL that carries only a containerized patient - no equipment, and no paramedics. That is a different service architecture, which is a hard sell for any over-regulated market, especially health.

Overall, a single-seater VTOL taxi is the most doable among all commercial VTOL applications in cities. Flying ambulances are a tiny market with requirements that are incompatible with cities - a commercial failure from the start.

As for EHang, they have no autonomy, a programmed flight automation at best, which means safety below Tesla autopilot level. When Tesla crashes into a parked police cruiser, it is a police report and a youtube story. When VTOL crashes into a populated area, it is a story that starts with immediate ban on operations, and ends with a billion-dollar settlement. That is how close they are to commercial success. As I said before, a proven autonomy is the condition for anything flying over populated areas to be a commercial success. I am not aware of anyone achieving or even claiming to achieve that yet. All promoted VTOL projects picture 'the future' with implied autonomy, but end with a reservation 'until then it will have a pilot'. An exception is EHang, which has some 'control center', meaning it is programmed or remotely operated at some degree, and the data on such mode of operation is already abundant from ~20 years of military flying much safer winged drones.

Actually, you don't need that much speed. hybrids can use their wings at any speed that's not zero, and of course they can still fly at zero. They shut down their rotors above their stall speed, You might not station the air ambulances at every hospital.

But yes, for ambulance, non-fixed wing designs can make more sense as efficiency and noise are not top priorities any more.

You don't think a loud ambulance would be tolerated in suburbs? Compared to the ones with sirens?

While I think you can get away on many emergencies with one paramedic (if patient can walk) but yes, you need to handle two.

As for automation, while you can have a pilot in an ambulance (not a full helicopter pilot, but one able to fly a semi-automated vehicle with skills just above drone pilot) the nice thing is that automation in the air is much easier than on the road. Really, only landing and take-off out in the field need a human, and only landing if the take-off area is clear of obstacles. Which is not nothing but it's nothing like driving. All the companies I talk to are not worried much on the autonomy front. They worry about these other issues like range, power and noise.

Use of authority agains population conditioned for obedience has its limits. Ambulances with sirens at 3am on an empty road, hospital helicopters hovering at 3am over a city, fire trucks with even louder sirens rushing to another false alarm - at some point that limit is reached, and it is believable that new traffic at low altitudes will reach that limit, as higher-pitched narrow spectrum of sirens is better stopped by structures that noise-like spectrum of air stream, or the low-pitch and infrasound from vortex of a helicopter. Remember, there was a time when supersonic passenger planes flew over land, were designed for that, and no one objected, and then they were banned - to this day. And all they did was making a not so loud boom now and then. It would be most unfortunate if urban air mobility met a similar fate. I personally consider that turn of events unacceptable, which is why 50, 60 or whatever dB value is within existing urban nose, is my personal limit. Prototypes may be louder, as it is implied that quieting can and should be done after the basic functionality is demonstrated, as it was done with commercial airliners. But quieting should be considered from the start, as EHang with its open rotors, ~20 minutes flight time and minimal carrying capacity clearly was not.

Automation is easier in the air, but automation does not provide functional safety in the air. The requirement is functional safety, not an illusion of safety based on illusion of control. Yes, it is not a problem to automate piloting on the level of "touch the map to go there", and it may well happen, until one day it does not. There was a time when risk taking was acceptable in civil aviation, crashes were common, and the public accepted that. That time is long past. Now the expectation is unconditional safety, something about 10-7 probability of fatality per flight; catastrophic failure probability is at 10-9 level, which is essentially 'never in the life cycle of aircraft design'. Automation cannot deliver that because of inescapable edge cases. Pilots get away with that, as they pay for mistakes with their own lives - that is some social contract that still holds. Once autonomy is demonstrated at sufficient level of functionality, the expectation will go up instantly. And remember the fate of the nuclear power industry: despite its best safety among others, one catastrophic failure killed the future of entire industry. Same thing can easily happen when vehicles start dropping on cities because of deficient design, or low quality code, or for reasons unknown. Civil aviation is still damaged by events of 2001 - just one catastrophic event almost 20 years ago.

Companies may not worry about autonomy, yes. As Tesla does not worry about its cars crashing into fire trucks, police cruisers, trucks that are too white, intersections that are too something, and so on. They may not understand what it took to reach that 10-7 safety level, and mistake urban air mobility VTOL for a flying gadget. I am sure Airbus does not share that attitude, and they are the ones dominating air mobility markets (after Boing crapped out).

A quick image search for Ehang I am not surprised it's noisy contra rotating blades in each overs airstream always are. If instead of eight fans on four arms it had eight arms and ducted the fans it would be a lot quieter and more efficient.

EHang is first to get one going, not the best design. In general the pure multirotor designs are going to be less efficient, but easier to make.

automation in the air is much easier than on the road

Seems reasonable at first glance, but why would that be true? Is it just because airspace tends to be less tightly packed? If so, that won't be true once these things become legal.

Maybe in a Class B airspace you can keep other traffic away. But outside of highly regulated airspace, there are an awful lot of issues mixing robots with live human pilots. And don't forget the birds. And the occasional balloon or other unmanned and uncontrolled flying objects. And below the restricted space (and sometimes above it), recreational drones.

But even then, what do you do when there's an emergency that forces you to land somewhere other than a runway? Can a robot really do that without endangering people on the ground?

(If it's just for limited purposes like ambulances, you can get away with just temporarily restricting the airspace. But then, for these purposes you just train at least one of the paramedics as a pilot.)

I think one of the reasons it sounds reasonable at first glance to say that flight automation is easier than car automation is that I tend to forget all the air traffic that isn't dealing with ATC.

Today, the airspace is not packed and automation is super easy. In the future, the airspace might get more densely packed, but it will still be nowhere even remotely close to how roads on the ground are packed. There will be only other aircraft and birds, no pedestrians, cyclists etc. We will eventually get lanes in the sky, but there is room for many thousands of them. VTOL computer flown aircraft can keep in their lane jut fine. If they truly need to, they can make sharp right turns, though this is wasteful of energy. So you'll space them more than cars, but again, where you have one lane for cars you can have 300 lanes for aircraft of this type. (Traditional human piloted fixed wing need more space in all directions and will have to keep to their airspace.)

There are many different visions of ATC in this world. Some plan for transponders in every aircraft (I don't think that's wise.) Some want real time communications with ATC computers. I like something simpler. Before takeoff, you ask ATC for a path to your destination. It draws you one in 3D that intersects no other path, and which has a variety of emergency exit paths all along the route which also intersect nobody (unless somebody else is in emergency descent as well, but you have a protocol for that.) You don't have to talk to ATC for the rest of the flight. You do need high accuracy localization but you get that from a variety of sources. (You also have a plan for something denying all the GPS and local localization beacons for everybody to get to ground quickly.) You only need to talk to ATC if you want to change plans, and ideally in an emergency. But if you deviate, everybody will see that on their sensors and tell ATC anyway.

It seems that you also are forgetting that not all traffic is controlled by ATC (really, despite the name air traffic "control," none of even an IFR flight is "controlled" by ATC). A tremendous amount of traffic is VFR. There may not be pedestrians, cyclists, etc., but there are birds, drones, hot air balloons, skydivers, etc. And there are weather conditions that make flight unsafe, which is a much bigger problem for aircraft than for ground vehicles (ground vehicles can usually just slow down in almost any weather; aircraft often have to deviate around weather).

Maybe one day in the very far future we'll have a world where you can't fly an aircraft without getting a clearance from ATC, just like maybe one day you won't be able to use the roads without getting a clearance from GTC (ground traffic control), but that's not the world we currently live in.

In the world we currently live in, at least outside of the most tightly restricted airspaces, pilots have to deal with VFR traffic and weather in addition to taking directions from ATC.

(Drawing out a 3D flight plan that intercepts no other path is already done for an IFR flight. Unless ATC can perfectly predict all future weather, deviations are not just necessary for emergency situations, they are incredibly common. So for an IFR flight, you're always going to have to "talk" to ATC along the way. Flight paths get rerouted during the flight. It happens, often, no matter how well you plan. Real time communications between ATC computers and aircraft computers could work, but only if VFR traffic is not too heavy. That's equivalent to saying that 100% autonomous traffic on the roads would solve autonomous driving. Sure, it would, but that's not the problem. The problem is mixing humans and computers on the same roads or in the same airspace. Switching all flights, commercial and GA, to centralized computer control, would be an expensive, enormous undertaking, essentially equivalent to switching all road traffic to centralized, computer control. Even better, aircraft to aircraft communication could be used with decentralized control. An advanced and automated version of TCAS would do wonders for traffic control, but TCAS is already expensive even without adding all the bells and whistles you'd want for automating traffic.)

Even your example of an air ambulance would be a good reason why deviations from a prefiled flight plan would be required mid-flight. The air ambulance could file a flight plan right before each takeoff, but you'd have to reroute all the conflicting flights around it. Automated digital communication from ATC computers to aircraft computers would be the way to go, but these air ambulances are not going to be operating solely in Class A or Class B airspace, so that doesn't solve the problem of maintaining separation from VFR flights.

Where would they land, anyway? How much space is needed for landing? How would patients be transported from the scene of the crash to the eVTOL landing zone? Would fire fighters have to be sent to the scene to establish the eVTOL landing zone, like they do with helicopter landing zones? Do they just land wherever and hope everyone gets out of the way?

Yes, the airspace is relatively empty right now compared to roadways. That's because traveling by air is very expensive compared to traveling by ground. If you want to change that, you're going to have to deal with the explosion in traffic that would result. Personally, I don't see it happening.

The plans for advanced systems to handle hundreds of thousands of simultaneous flights are only for the dense urban areas. And yes, old school aircraft will need to keep to their allocated airspace if they want to fly the old way, and that airspace will shrink. They won't be able to just do a VFR trip over a city.

This is just one model. Most of the other models being worked on do expect regular communications with ATC systems. (Not people.) I am wary of those, but everybody knows you want a system that survives the failure of any central computers gracefully. We'll see which approach wins.

The air ambulance doesn't have to do major rerouting of traffic any more than it does on the ground, by sounding a siren. There is tons of room in the sky, enough that you can even leave flight levels for only emergency and diverted traffic and get away with it. Our ground network up the SF peninsula has 2 highways and 2 major routes, all at ground level. In the air are 100 routes at 50 flight levels, but that's not counting the ability to fly over the bay, or to do it when crossing the bay which you today must do on 3 bridges, where again there are 500 lanes and 50 flight levels.

There's room for a lot of traffic, and a lot of emergency and change of plan traffic too. The main issue is true emergencies where you've lost all power. That should be pretty hard to happen in vehicles with 8 rotors powered by 8 power systems which can do an emergency landing losing 3 of them, electric motors that effectively never fail.

And yes, old school aircraft will need to keep to their allocated airspace if they want to fly the old way, and that airspace will shrink.

If that's necessary, then automated flying isn't easier than automated driving, you're just proposing a much more infrastructure-intensive solution.

If we could keep human drivers off the roadways, autonomous driving would be pretty easy by now.

The main issue is true emergencies where you've lost all power. That should be pretty hard to happen in vehicles with 8 rotors powered by 8 power systems which can do an emergency landing losing 3 of them, electric motors that effectively never fail.

Electric definitely has a big advantage over avgas and jet fuel on that front.

We'll see how quickly the FAA goes about approving these things, though.

I don't see why slowly taking away airspace from the old school (which currently owns all the airspace at certain flight levels and zones) is a more infrastructure intensive solution. What infra are you talking about? I don't think we need to take that much away from them, but I just suspect we will as they become more advanced (right now they use immense amounts of airspace with huge spacing because they need to.)

What do you mean when you say the "old school" "currently owns all the airspace at certain flight levels and zones"? I'm not sure what you're referring to there. I'm also not sure what you mean by "slowly taking away airspace." Nor by "they use immense amounts of airspace with huge spacing because they need to." Can you be more specific?

Does "old school" refer to VFR? Does "huge spacing" refer to the 500 feet of separation required by VFR aircraft?

I can't even guess what you mean when you suggest that any airspace is "owned" by "old school." The old school concept is that airspace is not owned. Anyone with a license can use it as long as you follow the rules. Pretty much like most roadways, though without the traffic lights.

Currently, ATC desires large separation between aircraft. Well, of course you always want as much as you can get within reason, but current aircraft are viewed as needing it and many of them do. Separation in flight levels. In lateral. In longitudinal. While skilled pilots can fly in formation, this is not done. Wakes from large aircraft require long clearances behind them. Automated, electric multirotor aircraft can get much closer safely.

Zones around airfields are "owned" or referred to as controlled. You don't fly in controlled airspace without informing and receiving clearance from the center controlling that airspace.

It is much easier for an ai to recognise obstacles hanging in mid-air than on the cluttered ground environment. With 360 unblinking sensors and unlimited attention span it will likely be much better than human pilots with their two measly eyeballs at that task. The ai will also be much better and ranging the object and plotting its 3d path and potential collision course, for a computer that is simple trigonometry. If it has the kind of radar sensors a Tesla has it can even do this in fog. Finding a piece of ground flat enough with no obstacles for an emergency landing does not seem that hard a task for ai either. There is not much extra magic a human pilot can do on that score, if its a real emergency he is relying on bystanders seeing that a helicopter is landing so they dont run into the middle of the field. A helicopter is noisier but an emergency landing system would likely have some loud hailer warning.

I suppose you're right when it comes to avoiding traffic in the air.

A forced landing somewhere other than a dedicated landing zone might be a problem, though.

Crash landing anywhere is bad news. For emergency landings where you still have enough rotors to descend, they will have a map of 100,000 parking lots and streets and driveways and I guess even rooftops. Indeed, if they can only take off and land from special approved spots, those spots won't be very far apart and will be kept clear (unlike random parking lots which of course might have cars parked in them.)

I do worry about what to do in an "everybody out of the sky now" event. Though short of military attack it's not clear what would cause one with no warning. Possibly terrorism.

Maps can't tell you where it is safe to land right now.

No, maps can tell you places it is likely to be clear, graded on how good they are and how likely it is they will be clear, and you can use those to guide your descent and then see what's clear when you get lower. As a plus, VTOL aircraft can land in places you should never land outside of an emergency, with low risk of injury, though possibly with property damage. I mean you could land on top of cars in parking lots or fields of grass if you had to. Even the middle of busy streets with flashing lights and a siren.

Key to this is that full failure of an e-VTOL should be almost unheard of. Most designs will be able to lose multiple rotors and still fly for a while. Lose even more and still descend at a rate that won't kill you. With multiple battery packs of which several can fail.

Many of the designs also have fixed wings, and can fly to and land on a short airstrip with just one forward rotor working. Electric motor failure is very rare.

Of course there can be catastrophic failure, main structure failure, EMP blast perhaps. Nothing will be perfect, but it should be very good.

Currently, ATC desires large separation between aircraft.

ATC governs separation of IFR flights, not VFR flights. Under VFR, pilots are responsible for maintaining 500 feet of separation.

Zones around airfields are "owned" or referred to as controlled.

I can see how you could refer to Class B airspace as "owned." Is ATC what you meant by "old school"?

Most airports are not surrounded by Class B airspace. In fact, the vast majority of airports don't even have services provided by ATC. "In fact, nearly 20,000 airports in the United States are nontowered, compared to approximately 500 that have towers." "As of August 2017, there are 37 Class B airports in the United States."

You don't fly in controlled airspace without informing and receiving clearance from the center controlling that airspace.

You need clearance to enter Class A or Class B airspace. You don't need clearance to enter other classes of airspace.

Who is it that you're suggesting that we take airspace away from? It's not Class A or Class B airspace that is going to be the biggest problem giving to robots. That airspace is already highly controlled. It's everything else (the vast majority of airspace below 18,000 feet) that's an issue.

As I said several messages up, "It seems that you also are forgetting that not all traffic is controlled by ATC"

Right, it's VFR flying in uncontrolled but urban space that is likely to need to change, probably with transponders in those aircraft, and systems which warn them if they are going where other traffic will be when they get there.

That's kind of like saying that self-driving cars are easy it's just human driving outside of limited access highways that has to change.

I think it's likely that we'll build roboaircraft that can fly under VFR in uncontrolled airspace before we get rid of VFR in uncontrolled airspace. Amazon already has approval for remote piloted drones test flying under VFR in uncontrolled and controlled airspace. To get that approval they probably demonstrated quite a bit of autonomy, as they had to prove the aircraft was safe in the event of a loss of communication with the mothership.

(Although the term "uncontrolled" usually means something else, here I'm using it, as you seem to have used it, to refer to anything below Class B, i.e. any airspace where aircraft do not need clearance to enter.)

Add new comment