Downsides to Robocars


For part five of my series on Robocars, it's time to understand how this is not simply a utopian future. Consider now:

The Downsides of Robocars

Every good technology has unintended consequences and downsides. Here I outline a few, but there will be more than nobody sees today. I still judge the immense upsides to be worth it, but you can judge yourself.


A downside that you fail to mention is that self-navigating vehicles are a necessary component of robotic cavalry. Strapping weapons and targeting systems onto these vehicles creates a robotic war machine with little risk of military personnel casualties, and a large possibility for abuse by unscrupulous dictators and other tyrants. Imagine if you didn't need to convince your soldiers of anything, including who the enemies of the state are. These cold-calculating warbots would kill anyone they were ordered to, even the citizens of their own country and unarmed civilians.

Also, imagine a car-bomb that delivered itself. I can imagine it would be a while before any of these are allowed in the West Bank or Gaza Strip.

Actually, I do have a section on that, also blogged it years ago. It might be in the roadblocks section, in that the government will fear robocars because they can be weaponized (while building bomb-bots for the military at the same time.)

I think your overall approach to categorizing the future possibilities is excellent. However, I think you have underestimated a couple of downsides. First, the fear of loss of privacy and increased govt control will be massive, and there will be plenty of people who will play upon that fear for their own ends.

At present, there are many areas of the country where the majority (in some cases the vast majority) of drivers are breaking the law on a regular basis (see LA freeways whenever they aren't a parking lot). This is usually, but not only, the speed limit. The idea that a car will not go faster than the speed limit will irritate a large portion of the population, especially given the arbitrary and often ignorant way speed limits are set today (far too low, often with revenue generation in mind, or in the false belief that it will significantly alter safety statistics). Without a vastly more realistic regulation policy for the roadways, law abiding robocars will be shunned by many.

The idea of being tracked precisely by your car's communications with the road network will also scare those curious enough to look into it. One only need look at OBD3 and GM's work on car to car networking to see the near term direction of this path. I agree that people are already traced and observed far more than they realize (credit cards, cell phones, cameras, etc.), but many are not conscious of it. I think it will be much more difficult to minimize that concern with the road network and vehicles you envision.

In the end, I think a significant proportion of the population will eventually adopt such vehicles for convenience, cost reasons (such as higher legislated fees for non-robo vehicles - I'm sure that many govts will use that as a "nudge" to get people into more easily controlled transport), and other factors. Let's face it, most consumers are very herd oriented (sheep?). But barring the legislation out of existence of human driven vehicles, a significant portion of the population will want to continue driving themselves without outside inteference.

And thus today I released an expanded sidebar article on privacy issues to outline some ideas here.

Realize that cars which communicate with the road network, directly or indirectly, are going to be coming with or without robotic driving. The robocars I advocate are autonomous, and receive information mostly, only transmitting requests to the city for reservations to take space in congested areas or get stopping space. With congestion pricing, we're already seeing cities track all cars and where they go in city cores. And sadly, we are all carrying mobile phones that track our every move, now to GPS accuracy in many cases.

I am also of mixed opinion on the traffic law questions. Part of me hopes that there simply won't be the need for as many silly traffic laws -- if vehicles are super-safe, you don't need to limit their speed -- but of course due to liability designers will be conservative. And that's only after humans are off the road which is a long way away.

Not only will parents be willing to consider schools that are far away from home, robocars would enable people to choose schools that are in the opposite direction of their commute.

Consider this hypothetical situation: Single mom has job requiring her presence from 8 am to 4:30 pm. Child's school starts at 8:30 (and opens its gates at 8:00,) and ends classes at 3:30 (and closes the gates at 4:00). If mom takes the child with her in her robocar to her office, and then sends the robocar back to the suburbs with the child in it, she eliminates the need for before school child care. If she sends the robocar to pick up the child at 3:45 or so, and drive in circles for a while before picking her up at the office, she eliminates the need for after school childcare.

If robocars come to pass in the same era as statewide school vouchers, many people with school-aged children will buy a robocar just for the children, and send them to a school on the opposite end of town in order to have reliable low-cost child care. The more children you have, the bigger the savings. My children attended public school for two years, and the school had no after-school care. I had to pay a fortune to have the children picked up in a special bus and taken to the YMCA. And it was a nightmare when one or both children missed the bus. There could be robobuses to drive whole neighborhoods full of kids around in circles.

Also, many people choose a home based on the home being within walking distance of the local junior high, the community swimming pool, tennis courts, public library and ice rink, so that they won't be enslaved by the child's schedule, or so that the child can be involved in lots of activities despite the fact that both parents work. With the robocar system, every chubby kid in America will be chauffered everywhere.

When will the robocars be available?

Obviously there will be limits on how long you can leave a child unattended in a robocar, even one configured just for carrying a child, which has nothing to break, a TV for them to watch, and won't obey most of their commands. Unlike leaving them locked alone in a room in a house (which would presumably be illegal) they are being delivered at a set time to responsible adults.

It is also unknown whether the scenario I described -- where the robocar has a videoconferencing link to a responsible adult, who can supervise the child, and redirect the car to a nearby responsible adult in the event of problems -- would alter parental and social attitudes about this. It also depends a lot on the age of the child. Certainly in the "fairly responsible" zone, where they are allowed on their own for modest periods this should be OK. For children of day-care age it may be a different matter.

I find this one hard to predict, and society could respond in a lot of ways. Indeed, some parents will crave it, though I don't know if they would go so far as to make the child's trip deliberately longer like that. The child of course would be watching TV, playing computer games and videoconferencing with parents and friends during the trip, so they might be reasonably happy with it. If it's the main part of a fairly limited TV/internet budget, they might be very happy for the long trip.

Robocars because of their superiority will supplant virtually all other modes of short distance transportation? Thus, all the authorities have to do to limit anyone's access to any place will be to delete it's entry in the "motornet". Once a destination is not accessible to robocars, it will not be accessible, period.

In fact, if they wanted to, all the places you could go could be limited to small list. Of course that is an exaggeration, but what if there was an "emergency"?

You know geeks might have a libertarian bent and see only the rosy side of this, but you know society is full of control freaks who will use these things to institute the ultimate nanny-state. Even it doesn't happen in the US you know some serious bad stuff will be done with these things somewhere.

There is no technical reason for what you say -- the motornet I describe is entirely optional, as an assistance, but no more necessary to robocars driving than traffic lights and lane makers are to humans -- but there is certainly a danger that the government could try to apply such rules, and as we design the robocars we want to assure they are designed to preserve freedoms.

ctrl+f for "which its huge" and replace which -> with

A small nit but I think the energy vs. velocity is a cube function not squared as you mention in the article.

The drag force and the power needed to overcome it goes up with the cube of speed. However, the energy per mile only goes up as the square because you are going faster, of course.

Power and energy do get confused in folks' mind a lot.

I wanted to agree in part and differ in part with your assessment of the car industry. Zipcar, Hertz and others will do much damage to domestic demand without autonomous cars piling on. However, replacing our current fleet and international demand, growth and ability to utilize autonomous cars efficiently leads me to think that legacy carmakers will continue to grow for many years.

However, even if this is not the case, newcomers like Tesla Motors, which will integrate autonomous technology seamlessly, will grow as vehicle electrification, drive-by-wire, and inductive charging become commonplace.

Throw in lithium-air, which should hit just as autonomous vehicles become ubiquitous, ~2020 according to IBM and their 500 mile battery project, and you have a recipe for incredible change.

Thanks for providing such a great source of information.


Car share is cool but really is only taking off in certain segments as far as I know. The need to return the car to where you got it, and the requirement to walk to and from the car share lot are huge downsides compared to a vehicle that picks you up and drops you off with no other worries for you -- much more efficient in sharing too, since you don't have to park or "own" it while at your destination.

My more recent thinking suggests that we'll actually put more miles on cars than we do today, though we will need to own fewer of them as a society. So when you think of a car's depreciation as pricing it by the mile, instead of the year, the car industry does not shrink quite so much. It does change what it makes and how it sells it, though.

\I like the picture you paint. I would like living there. This is a very good thought exercise. It is generally well thought out except for some very poor assumptions.

The notion that sprawl is a downside is the most egregious. When you say cars spawned suburbs, your have it exactly backwards - basic human needs for space, peace, and quiet spawned suburbs; automobiles made it possible.

I have spent some years living in very dense, mixed-use urban areas. High density mixed-use development is horrible for quality of life. Those errands which take one far from home necessitate dealing with horrendous congestion due to the higher concentration of people in smaller spaces.

Visit Ho Chi Minh City in Vietnam if you think that robocars are going to reduce congestion. 90% of road traffic is motorbikes which are smaller than your proposed robo cars. They flow around each other just like your robotic utopia. The congestion is terrible at almost any hour of the day. Too many people and not enough space. Also, excellent infrastructure and public transit make little difference - visit Hong Kong.

Exurbs and Suburbs are far more sensible and human for many, if not most, families than urban centers. Downtown walkable areas are nice but should not be considered as some kind of ideal. They suit some folks, but many people legitimately hate living on top of one another and having to fight traffic to get to big-box stores and open spaces. Sprawl is a good thing.

Los Angeles is an excellent model of distributed urban cores with surrounding habitable spaces as well as natural public spaces. It's not perfect but it sure is nice in those regards. Robocars would suit LA very well.

The second poor assumption is the notion that green is an unequivocal good and that we want to get off fossil fuels. Fossil fuels are safe, ubiquitous and by far more efficient than any other politically or technologically viable option at this time (nuclear is positively the best alternative whose benefits I note you recognize). Cleaner things will come, but only when they are better and cheaper than fossil fuels. For now the best energy policy possible is to replace gasoline and coal with fracked CNG. The US economy will likely slog in the doldrums until this happens.

"Going green" should be looked at, not as some kind of environmental good, which is basically irrational for most people, but as an economically rational way to do things better, faster, and cheaper. Keep in mind that, as a rule, humans do not live long enough to benefit from long-term thinking and most of us lack the education and life experience to understand concepts like externalities and hidden costs in any meaningful way.

It should be understood that concept of "rational" when applied to economic human decision making really does not exist in any conventional sense of the word rational. My only economics class started with an explanation of how humans do not "need" to breathe. Those folks have a way with words - mostly a way of turning meaning on its head.

Economic terminology and the limits of human rationale notwithstanding, recycling is often more wasteful, costly, and pollutive than traditional waste management except where it presents a cheaper alternative compared to mining, refining, cartage and landfills. The environmental impact of a hybrid vehicle's gasoline savings are laughable when compared to the impact of the batteries.

The "clean" energy resources you so earnestly desire are not going to come from crony capitalist campaign donors getting kickbacks from client politicians in the current model. They are going to come from giant petrochemical corporations who finally realize that they are in the energy business, not the oil business, and that investments in other kinds of energy sources (which happen to be cleaner) are going to make them more money with fewer headaches.

Robocars offer tremendous value propositions and efficiencies. "Green" benefits are a byproduct of those efficiencies. To make them the goal puts the cart before the horse. While there are a number of people who embrace "green" for it's own sake, the way to sell robocars to the general public is by showing us how they make our lives better and save us money. It's an easy sell.

I'm totally on board with robocars. I haven't owned a car in years, because it doesn't make sense for me. I've had a private car with a driver in the past and that was the ultimate luxury - bring on the robocars please. Having my own driver was the best way to get around ever, especially compared to public transit, cars, motorcycles, taxis, bicycles and walking - all of which I've used as my primary method of transportation.

Also, anything which rids the world of "professional drivers" will be embraced and welcomed by the whole of the world except those who make their living as truckers, taxi drivers and teamsters. Making them starve wouldn't be as quite much fun as watching them burn in a fire but at least we wouldn't have to rely on them any more.

I exaggerate. These folks have a tough enough life as it is. But they should not expect much sympathy from the general public. Theft and graft are far more memorable than good experiences. Folks basically want these jobs to be done as if by robots. There is a broad exception for those arriving in a new town and wanting some personal advice on where to find illicit entertainments. Jobs driving taxis are not going to disappear.

You've made some other poor assumptions - look for non-economic value judgements and try to understand how silly they seem to folks who do not share your values. Regardless, what you postulate has a great deal of validity and shows careful consideration. Incorporate an understanding of how many people think the world actually works and you'd really have something.

One downside you've omitted is that your vision is unrealistic and incompatible with cultural inertia. In the US we have entrenched notions of autonomous personal transportation, a trend towards metastasizing government and a bloated legal profession. As such, the robocar technology and will be ready long before the the laws will and the laws we get will make things worse, not better. Basic questions of liability will make lawyers on both sides plenty of money with no benefit to the public and will result in monstrous legislation.

Politicians, particularly on the left, will fight to destroy the potential of and kill the market for these devices by insisting that they serve some client interest. Folks on the right will be up in arms at the mere mention of roads where non-robotic vehicles are not permitted, never mind whole municipalities. There will be a not insignificant number of folks in the general public who will resent the loss of autonomy or jobs.

I come from a country which is unfortunately governed by folks like Barack Obama, Harry Reid and Nancy Pelosi who never met a good idea they didn't oppose and folks like George W. Bush and Mitt Romney who can't embrace a good idea until they find a way to screw it up.

Again, I exaggerate, but basically that's what this vision is up against. Except that we'll have to get these kinds of folks to agree on a compromise which will doubtlessly be worse than worst things they come up with by themselves. Add the client groups these people truly serve instead of the public into the mix and folks like you and I who would rather that driving was done by robots are sure to be sorely disappointed.

Finally, what you have to say about parking and RV's is patently absurd. It undermines every good point you make so fix it or lose it.. Robotic operation is very unlikely to change the basic value proposition of RV's. Price buying one and finding a place to keep it. With regards to parking, spend an hour watching any sizable valet parking operation and thirty minutes researching the cost of "free" parking and figure it out.

I agree with you more than I disagree, which is pretty unusual considering you can talk about going green with a straight face. Lip service to green goes out the window when the time comes for us to open our wallets and there isn't enough money inside. Luckily for the planet the value propositions, quality of life benefits and cost savings of robocars will save most people time, money, and trouble. That they are going to truly care about when the bill comes.

Anyway, thanks for a fun read. I will welcome robocars with great joy.

Came to this article via "New design factors for Robot Cars" and to there via a comment on NextBigFuture...

The image of rich people with RVs for their office commute, or a small fleet of specialised vehicles following them around, made the SF-geek in me want to project that what-if out onto a whole society. If we had robot cars and some kind of free/cheap and compact energy source/storage, what-if society adopted "automated vehicle rooms" as an alternative to housing. Instead of homes, you own a fleet of robot vans that each serves the function of a single room; sleeper cars, bathroom-car, kitchen, dining and lounge cars, etc, as well as storage cars.

This could develop naturally. If wealthy business commuters did use ever larger robo-RVs as kitchen/bathrooms - crawl out of bed and straight into the robo-RV, breakfast and shower while travelling to work - then these single-function robo-RVs will trickle down to the poor. In our world it doesn't matter how luxurious a new car is, eventually even a Cadillac is owned by some guy who's so poor he has to live in it. And since the robo-RV essentially is a room... it makes economic sense to live in them. Something between a trailer-park and tenements would become the standard low-income housing. Then gradually be adopted by everyone else not just as extensions of homes, but as homes themselves. And because you only "live" in a couple of rooms of your house at any given time, then even if you still want a private block, you need less land for the house itself. The only permanent part of the "house" might be a road-backed hallway, with docking doors for the robo-rooms to park against, and perhaps one permanent toilet. The rest of your block is all yard, with your unused robo-rooms stored in vast out-of-city yards that most people will never see and rarely think of.

So, you "live" in whatever module/room you want at the time, whether parked at your "home" plot, or at a temporary parking area, or on the road; summoning or scheduling whichever "room" you want next, to meet you wherever is convenient. Similarly, one set of robo-rooms but multiple plots of land; you live in one area during the week (when you want quiet at night, and/or to be close to work/school/shops during the day) another area on the weekends, preferring different groups of neighbours for different times of the day/week/year/mood. You could easily lose track of where you are at any particular moment, only concerned with having the right function at the right time (no matter where you are), or being delivered to the right location at the right time (no matter what "room" you were in.) Kids don't really "move out", they just get more of their own personal robo-rooms, and roam further as they get older, until one day they've assembled their own house and generally live in another city. Some large multi-generational families would end up in a single "home" sprawled across multiple cities and surrounding countryside, through many shared robo-rooms across the various family-owned "yards" and parking areas.

Even businesses might be mobile, with offices or even shops moving around to adapt to the people who use them. "Technology Parks" or "Business Parks" would be similar to "Trailer Parks", momentary collections of mobile modules. Cities and suburbs, and surrounding towns, smearing out and leaving whatever permanent infrastructure is too big/heavy to be allowed on the road, mixed amongst parks and automated farmland. And roads. Big wide roads.

Okay, the pragmatist in me knows that this won't happen. Hell the expense & maintenance alone, due to complexity, would outweigh the pure awesomeness. Kind of like how boats are always more expensive & higher maintenance than similar sized cars, RVs or houses. But still...

I hope the list isn't presented in your perception of magnitude (or even the order they occurred to you). Job loss is huge factor that should not be overlooked. Don't look for pushback and repercussions to be limited to organized protest. Look for sabotage, mob attacks, bombings, sniper attacks, etc., not to mention cyber sabotage. The world already has a HUGE employment problem and new symbols for labor displacement will be targets.

One missed factor is the individual effect of higher transportation costs. Those who cannot afford to transition will eventually be discriminated against (parking, insurance, vehicle licensing, tolls, etc.). On top of that transitional (20+ year) effect, you would be completely naive if you think the per mile cost of transportation will be lower in the new world than in the current one. A majority of the population will feel real pain if transportation takes a substantially bigger slice of our stretched budgets (I will).

Lots of real benefits. Huge amounts of infrastructural issues. Huge amount of new regulatory considerations. Probable inevitability, but probably also further out and slower transition than anyone predicts. Finally, real pain at the individual level that should not be forgotten.

Yes, I firmly believe the cost of transportation will go down significantly, for all.

Cars will be shared (particularly by the poor) and get far higher utilization. The typical car today costs a lot of money and sits idle 93% of the time. Robot transportation is way more efficient and saves a lot of money. The ability to use more efficient vehicles cuts fuel costs and saves even more.

Today people spend around $8,000/year to own a car. I expect them to pay less than half that in the future to get even greater benefits. Perhaps even a third.

Yes, people who drive for a living will be disrupted. But think about all the other people in the world who also drive for a living and don't want to, because they commute 1-2 hours a day. They don't drive 8 hours a day like the pro driver, but there are 50 times as many of them, and when they stop driving to make a living, they will turn that time into useful time for themselves, or make money during it. Should we really prevent that to save those jobs?

Exhaustive as your list is, it overlooks something fundamental — a driver's ability to make spontaneous decisions during a trip.

When I drive, I can choose which course I take. I can alter my destination en route. I can avoid roads I dislike. I can select a particular parking spot. If I see something interesting, I can stop for a photo (I'm an architectural historian, so I do this a lot). Heck, I can even drive my car into a random field, if I so choose. Most importantly, I can drive without a destination in mind.

Unless autonomous vehicles offer the sort of flexibility, control, and spontaneity that comes with conventional driving, I want nothing to do with them. To be trapped in a steel-and-glass coffin — to be whisked between points A and B without a modicum of decision-making power — would be a nightmare. It would be tyranny, and I'd refuse to submit to it.

Many people want that, so I don't see why vendors won't offer it. It may come at a very slightly higher price in some cases. If you asked for a vehicle to take you 5 miles and you suddenly say, "take me 200 miles" they probably will have it pause along the way to switch vehicles for you.

One thing that will be a problem is stuff like driving into a field. It may be harder to order a car ready to do that. However, for a very long time it will be possible to order up cars with steering wheels which drive you to the edge of the field where you do what you want. The problem I see is many rental cars today forbid going off road but have no way to enforce that. But some companies will exist for the off-roader and the nice thing is you can switch cars any time in the service areas of the providers. Extra cost, but still cheaper than today.

Add new comment