The flying car revolution will never happen

One of science fiction’s greatest mistakes is the flying car. And that icon of the future is under further threat from a real technological revolution, the driverless car.

My mental image of a flying car is shaped heavily by their depiction in the Fifth Element, in which Bruce Willis drives a yellow taxi that is, cosmetically, an ordinary car that happens to drive on thin air. He navigates traffic lanes that form a 3D lattice amongst buildings so tall that we rarely see the ground. A close second in my mental imagery is the “hover converted” DeLorean1 from Back to the Future part II, which we see navigating a “skyway” in the far-off year of 2015.

It didn’t happen, of course. There are certainly people trying to make it happen, but clearly they haven’t made much of a dent in the way our transport systems actually operate. Articles on the subject of flying cars (e.g. this one, and this one) often start by pointing out the obvious, and end by wistfully suggesting “maybe some day”.

No. Here’s my prediction: the flying car revolution will never happen. Not in a decade, nor a century, nor a millennium. The flying car — as portrayed in science fiction, and distinct from today’s commercial passenger aircraft — is a fatally flawed concept. That’s not to say you can’t make one; you just can’t make society adopt them. They will never be a significant component of our transport network.

Safety

There are several reasons. The first to spring to mind first is safety. Surface driving is already a significant cause of death; drivers already fail at the delicate art of steering in two dimensions, let alone three. When we do travel by air, we appreciate that our pilots are somewhat more qualified than your average tailgating, drink-driving arseclown who has no understanding of risk. And even when you’re not a tailgating, drink-driving arseclown, driving is still inherently risky.

Would you like to see what a flying car collision looks like? Observe from afar, because there’s no such thing as a “fender bender” in the air. There’s no “five star” safety rating. Seat belts and airbags won’t save you. If you lose control, you’re simply going to die. And in a populated area, people around you are going to die too, not to mention the property damage.

Some quick risk analysis is sobering. The probability (0-1) of something going wrong is bound to be higher. The consequences (measured in dollars or lives) are certainly much worse. We can multiply these to get a quantity called “risk” (also in dollars or lives), which you can think of as the expected cost per flying car, on average, if many people owned them. Since both factors are much higher with flying cars, the risk is catastrophic.

Security

And calculating the probability of an “accident” avoids another very serious problem: what about deliberate crashes? What about criminality, or terrorism? How can you even begin to police a “skyway”? The term “flying car” belies the fact that this thing isn’t just a car but a missile. Should we install automated air-defence turrets on parliament, banks, hospitals, schools, apartment blocks, etc.? We must. If we are not prepared to shoot down flying cars, then inevitably criminals and extremists will use them, whether in a kamikaze attack, or as a remote-piloted weapon. And their targets will always be high-value ones.

Perhaps manufacturers can design safeguards to prevent collisions, and/or prevent entry into unauthorised airspace. But that won’t stop anyone with an aptitude for taking things apart.

And if we are prepared to shoot down flying cars, would you get in one? Moreover, the air defences themselves might prove a tempting weapon for criminals and terrorists. Feeling safe yet?

Physics and fuel efficiency

Flying cars from science fiction generally rely on as-yet undiscovered physical phenomena to keep them airborne. Sometimes “anti-gravity” is mentioned, a force that simply doesn’t exist as far as the current pool of scientific evidence indicates.

Even if anti-gravity were to exist, it would not necessarily be of any practical use. Hypothesised “exotic matter” could have negative mass, ergo negative, repulsive gravity. But it’s Earth’s gravity that matters, not the car’s. The “acceleration due to gravity” near the Earth’s surface is the same for any falling object, irrespective of mass (if we ignore friction). An exotic object with negative mass, even if it were possible, would still just fall downwards like everything else.

Magnets? Well, no. Although magnetic levitation exists (see diamagnetism, as used in maglev trains) there’s a significant difference between levitating a few millimetres above a superconductor and actually flying. You would need an implausibly powerful magnet to hold a car-sized object tens of metres off the ground, and if you had such a magnet, it would probably destroy everything around it. You can’t just tailor a magnetic field to only apply to one specific object. And the ultimate reason for the total futility of this idea: you still need a surface-bound vehicle to hold the magnet (or superconductor) underneath your flying car. What would be the point?2

Obviously, we do have a few real ways to keep something airborne. Balloons use buoyancy, but they are large, slow, and difficult to manoeuvre. Rockets use a controlled explosion, but this lasts for a matter of seconds. The most plausible way to create an actual flying car is as a scaled-down plane or helicopter, with spinning blades and aerofoils to push down on the air.

But “anti-gravity” is used in science fiction to get around a huge problem with aircraft — their fuel consumption. In the real world, fighting gravity consumes energy. Continuously. A surface car needs energy only to overcome friction and inertia. A flying car needs large amounts of energy simply to stay up, on top of the energy needed to actually move it from point A to point B. And in the end, the point A and point B that you’re travelling between are both on the ground, or at least connected to it.

(A side note: there is the interesting phenomenon of electrohydrodynamics, which can create lift by making charged particles push air molecules, with no explosions or moving parts needed. But even this doesn’t get around the problem; it still requires very large energy input.)

So why are you burning up all that jet fuel3 just to keep above the surface, when you could be cruising along the surface at a fraction of the price?

Traffic congestion is perhaps the one genuine reason people might have for taking to the air. It’s certainly tempting to imagine going over the top of all those suckers stuck in gridlock on the bitumen.

Economics

But there’s a problem with this picture too. While you’re flying over all the suckers stuck in gridlock on the bitumen, it may occur to you to ask why they aren’t flying too. Consider this paradox: if everyone is flying, by definition there is no surface traffic, ergo no surface gridlock, so then why is anyone flying?

Flying cars are going to be inherently more expensive than their surface counterparts. If we do overcome all the safety and security issues, it’s going to be thanks to some pretty fantastic engineering. That costs an awful lot of money up front, which needs to be recovered in sales. Also, as mentioned, there’s the fuel. And then there’s the servicing, similarly expensive due to the extreme life-and-death importance of keeping the vehicle running. And speaking of extreme life-and-death matters, there are the insurance premiums, if someone is crazy enough to sell you insurance at all.

It all adds up to a fanciful luxury for the vast majority of people. It would be fantastic, we’d all say, right up until we see the number of zeros in the price tag. It would be fantastic, but it’s just not worth it.

Furthermore, getting back to the paradox above, the uptake of flying cars in itself discourages further uptake (the opposite of many other technologies). Every time a person chooses to fly, that’s one less surface vehicle, meaning less congestion. Every person who uses a flying car reduces the incentive for everyone else to do so.

Nonetheless, you might imagine a situation where (a) the population continues to increase more or less boundlessly, and (b) the government makes fewer and fewer investments in surface infrastructure in recognition of increasing numbers of people with flying cars. Perhaps everyone would use a flying car eventually.

But even given that we’ve somehow solved the safety and security issues, I still think this scenario is highly unlikely in any reasonably democratic society. Governments will spend money on infrastructure (whether roads or public transport) because these are tangible things that voters will easily understand and use. Infrastructure makes for compelling election promises, and although it’s a short-term hit to the budget, it’s often easy to justify in real economic terms. To abandon infrastructure spending would be to abandon large numbers of voters who can’t afford the costs of flying.

Indeed, in a hypothetical city that already relies on flying cars, there would be economic pressure to go the other way — to develop efficient public infrastructure in order to minimise costs. Large/dense metropolises tend to have extensive public transport networks, because as the population density goes up, the cost of public transport per person goes down. The cost of flying cars does not.

Automation

But even congestion and commute times may not be so much of an issue in the future. Driverless cars represent a rather different and much more realistic kind of revolution.

This innovation could dramatically improve the efficiency of our road networks. On the roads we have constraints: the speed limit, the distance from the car in front, the lane widths, etc. We need these because of our general human imprecision, our slow reaction times, our reliance on mostly a single sense for driving (our eyesight), our blind spots while driving, and our inability to make decisions based on large amounts of sensory information at once. It’s also essentially impossible for a human driver to see what’s happening along alternate routes.

Driverless cars, in principle, have none of these problems. They can be faster and safer and more efficient. They can, in principle, perform safely in situations that would be dangerous for a human driver. (I only say “in principle” because it’s a matter of engineering quality. It may take a while to develop driverless cars to this point, but there’s nothing fundamentally standing in the way.)

To do this, they can take advantage of sensory data that humans simply don’t have — sonar, radar and lidar — to model their surroundings in a level of detail many times beyond our own mental capabilities (which are really very limited in rapidly-changing situations).

Software and hardware can react in a tiny fraction of a second, so driverless cars can safely leave much less buffer space. They can also make the best decision on which route to take based on real-time measurements of traffic density (provided by some external source). As a result, we could squeeze a lot more of them onto existing roads, increasing the total number of trips possible with the same amount of infrastructure.

Many people are doubtless nervous about giving up control of their vehicles, but as discussed, human driving is hardly safe as it is. In making a driverless car safer than a human driver (safer even than you), the bar isn’t really that high. Safety will become one of the principal arguments in favour of driverless cars, not against them. It’s just a matter of design, testing and time. (I don’t doubt there will be accidents, but it’s the accident rate that we must focus on.)

Driverless cars also facilitate cheap, automated taxi fleets, which may allow a lot of people to get by without their own cars at all. No more parking, refuelling, servicing, insurance or licensing (all of which are taken care of for you). No more fines and tickets. No more traffic police, except perhaps for the few DIY holdouts. Unlike flying cars, automated taxis would be a self-reinforcing technology. There’s no huge up-front to commuters, and the more there are, the shorter your waiting times will be, so the more likely you will be to use them.

And when you’re commuting in a driverless car, that need not be wasted time. You could be working on a laptop/tablet, reading, teleconferencing, eating a meal or even just sleeping. Given that you can fill in the commute with useful activities, the actual time it takes may not be so much of an issue any more.

It’s not just that driverless cars are more useful and realistic than flying cars. Once we have driverless cars, the arguments for flying cars will practically vanish altogether.

So, science fiction, let’s put one of those ideas out of its misery.

  1. It’s also a time machine, but that’s strangely unimportant in this context. []
  2. Okay, perhaps you could convince the government to build a grid of super-powerful magnets to keep your flying car aloft, but I suspect some political advisers might prefer to aim for more efficient and less catastrophic infrastructure projects. []
  3. It could be bio-fuel or batteries, of course, so supply is not an unsurmountable long-term problem, but it will be expensive nonetheless. []