Month: March 2019

Ah, flying cars, a staple of science fiction since The Jetsons, how I hate thee. Let me count the ways…

First, let’s define what we’re talking about: A “flying car” is a vehicle that can fly through the air like an aircraft AND ALSO drive on roads like an ordinary car. Thus, though it might take off and land vertically like a helicopter, a flying car is different from a helicopter because it can also move long distances on the ground.

In theory, flying cars would be more versatile than land-only cars and air-only aircraft, but their dual-role nature would impose design compromises that would make them far less efficient than either of the other two. For example, a flying car’s wings would be useless dead weight and bulk when the vehicle was driving on roads, and its wheels and transmission would be useless dead weight and would produce major drag when the vehicle was flying through. As a general rule, flying cars would be heavier, slower and less fuel efficient in the air compared to small aircraft, and more prone to breakdowns, less safe, and less fuel efficient on the ground compared to normal cars. 

Without getting into any more detail, we can say that flying cars are a flawed concept, and there’s no reason why this shouldn’t have been obvious to engineers in the 1960s (or earlier) when The Jetsons aired and implanted in the popular consciousness the idea that flying cars would be common in the future. Unfortunately, none of those engineers spoke up (or maybe they did, but they were ignored), and flying cars went unchallenged. I think it’s unfortunate that so many works of science fiction featured flying cars, as they created an unattainable expectation in the minds of millions of people, which has led to predictable disappointment with the way things actually turned out and helped to prop up the false arguments of cynics and declinists. Peter Thiel’s famous quote aptly expresses this misguided disillusionment: “We wanted flying cars, instead we got 140 characters.”

I don’t like flying cars because their failure to appear by the deadlines set by works like Blade Runner is often held up as proof that technology is not improving and our lives aren’t getting better with time. As a student of history, I know that is badly wrong. I also don’t like them because they’re examples of bad futurism–They’re a future technology that sounds superficially cool, but that can also be shot full of holes by any reasonably smart person who spends a few minutes thinking about it critically, as I’ll now do in detail.

Using a thought experiment to build a hypothetical “flying car” from existing technology puts the problems in stark relief. Let’s start with a classic, reliable small plane–the two-seater Cessna 150–and mod it to be a flying car. The first problem we run into is that the wheels at the ends of their three landing gear aren’t connected to the engine by a transmission, meaning the pilot can’t make the wheels spin like he could in a car. Instead, pilots do ground taxiing by increasing the power to their engines, and the spinning of the propellers or jet blades pull the aircraft forward, just as they do when the plane is up in the air. Steering on the ground is done through differential braking of the wheels, and at higher ground speeds, through use of the rudder. While this is fine for traveling a few hundred meters from an airport hangar to a runway, it’s grossly unsuited for driving on roads with normal car traffic.

We have to add a transmission that connects the Cessna’s engine to at least one of the plane’s wheels, and we also have to add some kind of mechanism to the engine that can disconnect it from the propeller when the craft is in “ground mode.” After all, driving down a residential street with a loud, spinning propeller at the front of your vehicle is obviously unsafe to pedestrians and would violate noise ordinances. We also need to add a feature that makes the wings fold up at the push of a button so the plane can be narrow enough to drive on standard roads. Installing the transmission, disconnector, and swivel mechanism adds weight, cost, and mechanical complexity to the Cessna.

So now, we’re ready. You put your modded Cessna 150 into “ground mode” and take it out for a spin. After a few minutes, you realize it’s the worst car you’ve ever driven. Your engine is literally five times louder than the cars around you and you’re constantly getting stares and seeing pedestrians around you covering their ears. Your “flying car” handles worse than a loaded dump truck (poor acceleration, wide turning radius, very mushy steering), struggles to reach highway speeds, and gets awful mileage. Finally, its small wheels and lack of a suspension system ensure every pothole and small rock on the road jolts your spine up into the base of your skull.

Though the vehicle folds up its wings at the push of a button to make itself narrow enough for you to drive on the road, it can’t shorten its 24 foot length, which dwarfs massive road-only vehicles like Chevy Suburbans (ONLY 18.5 feet long) and gives you a huge turning radius. But paradoxically, your Cessna 150 flying car doesn’t have any more interior space than an ultra-compact Smart Car: There are just two front seats and enough cargo space in the back for a full load of groceries. The ride is cramped and uncomfortable, you can’t use the flying car to transport any kind of big cargo, like a piece of lumber from Home Depot that you need for a simple home improvement project, and it can’t be an all-purpose family vehicle if there are more than three people in your household.

And worse yet, when you decide to forget that stressful experience by switching the Cessna to “air mode” and taking to the skies for a fun ride, you notice the plane is much slower, less maneuverable, and can’t travel as far on the same amount of fuel as before. All the mods you added to the plane to make it better at driving on roads have weighed it down, and it suffers in flight. Other small planes designed exclusively for air travel zip by you.

If this sounds like a sucky thought experiment so far, realize it actually gets worse. Your modded Cessna 150 would need more mods to meet car safety laws, like airbags, bumpers, and crumple zones, all of which add more weight, cost, and complexity. Granted, if this thought experiment is set in the distant future and car accidents have become very rare thanks to autonomous drive systems, it’s possible that some safety feature laws will be eased or eliminated. But not all of them, and for sure your Cessna would need more mods.

And as a person with discerning tastes, you’d doubtless want to install bigger wheels and a suspension system under your craft so every drive to the local store didn’t feel like mountain biking over a jagged rock trail. Which means–you got it–even more weight, cost and complexity.

After finally transforming your super-modded Cessna 150 so it drives as well as a low-quality car, to your horror, you discover that it has become so heavy and non-aerodynamic that it can barely take off into the air anymore! Maybe it can’t fly at all! Uh-oh! And now to fix THAT problem, you have to do a totally different set of mods…and you see where this is going.

Put simply, aircraft and land vehicles have totally different sets of role requirements, and making a “flying car” that can do both forces major design compromises, and it will never be as good in either role as specialized craft. This is true regardless of whether the flying car has wings like a small plane, or rotors like a helicopter.

Speaking of that, I forgot to mention that another limitation of your modded Cessna is that it will only be able to take off from long runways. Unless you are part of the ~2% of the population that lives on a large plot of flat land in the countryside, this means you’ll have to drive to an airport every time you want to go flying. The extra time spent driving your Cessna flying car to and from airports will be an inconvenience, and will actually make it faster to use ground driving mode to travel short- and even mid-distances.

But if your flying car were instead based on a two-seat Robinson R22 helicopter, you’d be able to get around that problem and take off from your suburban backyard, or from the roof of your apartment building, right?

Kinda…maybe…sometimes.

This brings us to two very important but overlooked problems with VTOL-based flying cars: noise and downdraft. Helicopters are very loud, and it would violate noise ordinances and cause people hearing damage if helicopters routinely landed and took off in their neighborhoods. Helicopters can be made quieter by giving them things like exotic main rotor blades, and cowlings around their tail rotors, but these design features are very expensive and only reduce noise levels by a few percent. Rumors that the U.S. military has top-secret “silent helicopters” are unsubstantiated, and I doubt it’s even possible to make helicopters that are “quiet enough” to land in your suburban backyard without jolting your neighbors out of bed. If big chunks of spinning metal are slicing through the air at hundreds of miles per hour, it will make a lot of noise no matter what. 

But even if very quiet helicopters could be made, the next show-stopping problem is downwash. A helicopter is able to go up because its main rotors blow air down at the ground with enough force to overcome the force that gravity is exerting on the helicopter. During takeoffs and landings, when helicopters are flying low to the ground, the downwash can be strong enough to blow over nearby lawn furniture, break tree branches off, blow off roof shingles, kick up big clouds of dust from the ground, and blow small pieces of debris like pebbles around at high speed. The attendant risk of injuries and property damage will ensure that it stays illegal for people to have personal helipads in their suburban backyards. 

We can calculate an R22’s downwash by using this equation: 

The data for the R22 can be found here: https://en.wikipedia.org/wiki/Robinson_R22

In spite of the fact that our hypothetical R22 is modded with a transmission going to its wheels, an engine-rotor disconnector mechanism, auto-folding rotors, air bags, and all kinds of other stuff to make it roadworthy, I’ll be really nice and say that thanks to use of futuristic weight-saving materials, its overall mass (including passenger[s]) is just 1,200 lbs. That yields a downwash of 22.5 ft/sec, but unfortunately, it’s actually worse than that:

“Keep in mind that this speed [derived from the equation] is at the rotor disk. As the column of air is forced down below the rotor, it constricts, much like molasses being poured out of a pitcher does. In doing so, it reaches its maximum velocity at 1.5 — 2 rotor diameters below the disc.”
https://www.rotorandwing.com/2011/11/29/calculating-rotor-downwash-velocity/

So our “R22 flying car” produces a downwash of 45 ft/sec, which is 30.6 mph. That’s not hurricane-force, but it’s strong enough to kick up clouds of dust, blow common objects over, and hurl pebble-sized debris into a nearby bystander’s eye with enough force to send him to the hospital. If the approach route to your backyard helipad requires you to fly low over someone else’s house or any sort of public space, then the clock will be ticking on someone suing you. So unless you have a very large yard that you’re willing to build a helipad in the middle of, forget it. While we can debate what the pace and direction of technological and scientific development will be in the future, there is no debate that people will continue getting more litigious and fussy with time. Someone will sue you because your flying car is too loud, or because it hurt them by blowing debris at them (even if the claim is a lie).

Let me insert an important caveat, which I first noted in my Starship Troopers movie review: The noise and downwash of VTOL flying cars are only problematic if we assume they’re to be used in a future world full of humans. If, on the other hand, we assume the future will be populated by machines and not humans, then noise and downdraft won’t be obstacles at all since machines won’t have finicky senses or frail bodies that can get hurt by little pieces of high-velocity debris. It might also be possible to reduce some safety features in aircraft intended for machines that have bodies that are more durable than ours. However, it’s also likely that machines will be very rational and won’t have the same problems we do planning their actions in advance, so from a resource usage standpoint, they would rarely use flying cars as it would be too wasteful a means of transportation. Traditional vehicles like boats, railcars, and big trucks will remain cheaper ways to transport cargo than aircraft.

And if you’re wondering whether we could avoid these problems by inventing some kind of anti-gravity or gravity-cancelling device that flying cars could use to go up and down with blowing air at the ground or needing long runways, realize that such technology is impossible because it violates the laws of science. Our understanding of how the force of gravity works provides no avenue for it to be controlled in such ways (and even if it were possible, it might require impossibly large amounts of energy). If your craft is heavier-than-air, and if you want it to do controlled flight, you either 1) need to give it wings and an engine so it can take advantage of lift, or 2) need to give it a downward-facing fan or rocket nozzle to blow vapor down hard enough to overpower gravity. Those are the only options.

Finally, in “ground mode,” our “R22 flying car” would have the same inefficiencies and problems as the “Cessna 150 flying car,” such as poor performance and handling, excessive length but deficient interior space compared to ground-only vehicles, etc.

Another problem is that the standards for “airworthiness” are much more stringent than the standards for “roadworthiness,” so minor damage from something like backing your flying car into a concrete pillar in a parking garage, or having your side window broken by your neighbor’s kid throwing around a baseball in his yard will ground the vehicle until it is inspected and fixed. Flying cars would surely have advanced and extensive internal diagnostic systems to detect such problems, and they will refuse commands to take to the air if there were even a minuscule chance of in-flight mechanical failure. This means the autonomous drive systems would have to be almost totally perfect to ensure even the slightest accidents never happened. And even if that technology existed, you’d have no way to stop vandals or reckless people from disabling your flying car’s ability to fly by inflicting small amounts of damage on it. The availability of “flight mode” would not be reliable, and you’d always be at risk of getting stranded hundreds of miles from home after flying there and then suffering minor damage to the vehicle.

Bad weather will also keep flying cars grounded much of the time–just as is already the case for small aircraft–undercutting them as reliable means of daily transportation. Since piloting a small aircraft is very hard and dangerous, it’s unrealistic to expect a large fraction of the population to learn how to fly flying cars, so the vehicles will need to have advanced autopilot computers. For legal liability reasons, the computers would be programmed to fly very cautiously, and they would refuse to take off if there were even a small chance of hitting bad weather. Unless you are lucky enough to live in a part of the world with very mild, unchanging climate, this means your flying car will only be able to take to the air in fits and starts, preventing you from creating a daily lifestyle organized around the ability to fly from one place to another. This throws a monkey wrench into visions of a future where we all live on big estates in the countryside where land is cheap, and fly into the big city each day for work (also, why not just telework?).

Of course, even if you were assured of a safe landing, you probably wouldn’t want to fly a small aircraft through bad weather, since by virtue of their size, small planes and helicopters suffer worse turbulence than the big passenger planes most people fly on. Being flung around the inside of the cabin by every shifting gust of wind is upsetting for most people, and enduring that while also knowing your life is in the hands of a computer autopilot would be unbearable for a great many (this feeling of not being in control disproportionately frightens humans for complex psychological reasons). Most people can barely muster the courage to climb a ladder to clean their house gutters, let alone fly in a small aircraft. Fear of flying will be a big obstacle to flying cars, and an even bigger obstacle to flying motorcycles and personal jetpacks.

I’m still not done! Flying cars also make no sense for short-distance transportation, like moving around your own town or city. The extra time spent getting to cruising altitude and then landing would make it faster to just stay on the ground and use the roads. The fuel costs of vertical takeoffs and landings also would also be much too high to justify short-distance trips that could be done cheaper and (almost) as quickly with land-only vehicles. These problems both get worse if you assume lots of people in your town or city have flying cars, since that would lead to the equivalent of traffic jams in the sky, and you’d have to fly slower and hover while you waited for a helipad space to open at your destination.

Flying cars also wouldn’t make sense for long-distance transportation over intercontinental or even cross-continental distances, because their fuel tanks wouldn’t be big enough for the journey, and because taking a traditional passenger plane would be much cheaper and faster. Consider that the Boeing 787-900 at full 362-seat capacity gets 87 miles per gallon of fuel, per passenger (https://paullaherty.com/2012/05/25/boeing-737-vs-toyota-prius-this-might-surprise-you/). In comparison, a Cessna 150 gets about 44, and a Robinson R22 gets about 22 miles per gallon of fuel, per passenger. A Boeing 787-900 also flies at 560 mph while the Cessna 150 and R22 fly at 122 and 110 mph, respectively, so the big passenger plane will get you to your destination much faster.

This leaves sporadic mid-range travel, which I’ll define as trips between 100 and 400 miles in length, as the one transit niche where it might make sense to use a flying car. But how many people need to frequently travel such distances? If you live in a metro area (including suburbs and exurbs), you’ll be able to satisfy the vast majority of your recreational and social needs without having to travel more than 100 miles from home. And as I established earlier, if you work far from home, it would be a much better idea to telework from your house instead of flying to and from your office building every day (and in any case, at random intervals, bad weather would block you from flying to work, so you couldn’t rely on it).

Flying cars would definitely make it easier to take vacations to the farther-flung parts of your geographic region. As a resident of greater Washington, DC, if I had a flying car, I would go to New York City and the beach more often each summer since both would be quick day trips, negating the need to stay overnight and pay high hotel rates. I would also explore southeastern Canada, and go to my favorite Appalachian hiking spots more, but all of this would only translate into a few extra weekend trips per year. Like most adults, I have responsibilities that often keep me pinned down, and sometimes I’m just too lazy to leave town even when I technically could. If I had a flying car, most of the time I’d be using it in “ground mode” for short-distance trips, and would be griping over its poor performance, uncomfortable ride, and limited utility. I’d probably be better off saving money by just sticking to a ground-only car and accepting a reduced ability to go to New York and the beach.

The counterargument, which is “Just keep your normal car for everyday road travel, and buy a flying car for sporadic regional travel,” makes me realize that there is a different transit model that is better than the “one flying car per person” model shown in many sci-fi movies: What if we don’t build any flying cars at all, and instead build a dense network of airstrips and helipads that people could quickly and cheaply travel through using autonomous, rentable, “air-only” aircraft? What if we paired this with an autonomous carsharing model that would quickly move people to and from those helipads and airstrips? Such an arrangement would provide all the advantages of the “one flying car per person” model without any of the downsides.

For trips in and around your metro area, you would rent self-driving Uber cars that would stay on the ground. Since most (or all) of the other cars on the roads would also be autonomous, they would precisely coordinate traffic flows, meaning there would almost never be accidents or congestion. Cars would traverse the roads much faster than they do today. Additionally, since these vehicles would be designed solely for ground use, they would be optimized for that role and would be safe, fuel efficient, and comfortable inside.

If your job were far from your home, you would telework by using technologies that already exist, or, if that were inadequate for some reason, by using virtual reality technologies that will exist in the near future. The amount of energy required to power your teleworking equipment would be much less than what would be required to fly to your work site each day in a small aircraft or flying car, and if you teleworked, you wouldn’t lose any time at all commuting.

On the rare occasions when you wanted to go somewhere outside your metro area but within 400 miles–let’s say to meet with a very important client at the office building you normally telework to, or to take a weekend trip to the beach or a different city–you would have one of the self-driving Uber cars you normally use take you to the nearest airstrip or helipad. Assume this scenario is happening a few decades from now, and your country has invested money during the interim increasing the number and density of airstrips and helipads, so most of your citizens live within a 20-minute drive of one. They are typically sited just outside of towns or in industrial areas so no one is close enough to hear the sounds of the aircraft landing and taking off. It’s also very common for large buildings to have rooftop helipads.

Your self-driving Uber car takes you to the local helipad or airstrip, where you exit and walk a short distance to a waiting self-driving Uber helicopter or plane. Since the aircraft is a two-seater, and either you’re traveling alone or with only one other person, you don’t have to waste time going through a security check: You can’t take over the aircraft in flight since there are no manual controls and can’t do significant damage by blowing it up. The small aircraft flies you to the airstrip or helipad closest to your destination, and when you disembark, there is a second self-driving Uber car waiting for you nearby. Moreover, since the small aircraft is designed only for flight, it is totally optimized for that role, and is much more fuel efficient than a dual-role “flying car” would be.

Alternatively, we might use high-speed, autonomous Uber cars for 100 – 400 mile trips. The cars would be very streamlined and low to the ground for optimal performance at, say 100 mph. They wouldn’t be much slower than small aircraft for many journeys, and would be safer and possibly cheaper for passengers. If all of the cars on the roads were driven by machines networked to each other, then high-speed cars like this could safely share the roads with slower cars.

And finally, if you needed to quickly travel more than 400 miles, you would have a self-driving Uber car take you to the nearest big airport, where you’d disembark and go through the same process that exists today to board a large passenger plane.

In conclusion, I think flying cars are a flawed concept; it’s unfortunate that they’ve appeared so much in science fiction and created an unrealistic vision of the future for many people; and a transit model based around autonomous small aircraft, networks of helipads and small airstrips near population centers, and autonomous road-only vehicles ferrying people to and from the helipads and airstrips would be better than giving everyone a flying car. Moreover, I think the speed and efficiency of ground transportation could be greatly improved by autonomous cars, negating the need for flying cars to move people around cities that have bad road congestion today, and also opening the door to rapid ground transit across mid-distances. While flying cars and small aircraft can be redesigned to reduce their noise signatures (for instance, by using electric engines and installing helicopter tail rotor cowlings), it’s probably impossible to make them quiet enough to land and takeoff in densely populated areas without disturbing people to the point that they take legal action. I also think flying cars would be more feasible in world full of intelligent robots but no humans, but still wouldn’t replace older modes of transit.

Links:

1. http://www.cessna150152.com/faqs/performance.htm
2. https://paullaherty.com/2012/05/25/boeing-737-vs-toyota-prius-this-might-surprise-you/
3. https://www.osha.gov/SLTC/etools/hospital/heliport/heliport.html
4. https://www.rotorandwing.com/2012/02/01/leading-edge-quiet-please/
5. https://www.faa.gov/documentLibrary/media/Advisory_Circular/150_5390_2c.pdf