Will people with robocars hire them out as taxis when they are not using them?

Getaround lets people search for privately owned cars today

Update 2019: New research suggests the incremental cost per mile of electric robocars is low enough to alter some of the logic below. An update is in the works.

Will your private car hire itself out as a robotaxi when you're not using it? Back in my 2008 fictional story A week of robocars I was one of the first to talk about this idea. Indeed, I have written for some time that there will be 5 main types of car users in the robocar world

  1. Those who continue with their manually driven cars
  2. Those who own a robocar strictly for their private use.
  3. Those who own a private robocar, but have it hire itself out when not being used by its owner.
  4. Those who reduce private ownership (ie. go from 2 cars to 1) due to robotaxi use.
  5. Those who give up private ownership of cars and exclusively use robotaxis and transit.

Many people like #3, and Tesla has even declared it will operate such a network for Tesla owners will full self drive to hire out their cars. In 2009, a group of our students created car-sharing company "Getaround," partly due to inspiration from my 2006 proposal. Getaround lets people do short term rentals of private cars and recently got a $300 million investment round from Softbank.

In spite of that early excitement, I have grown to think that group 3 will be fairly small. It derives too much from old-style car thinking. Today we buy a car and, as many like to report, it sits idle up to 93% of the time. Why not take this big expensive asset and put it to work during all that idle time? This idea has even made some people imagine that somehow far fewer cars would need to be made if we could share them better. (It turns out there are far fewer vehicles parked, but more on the road and more manufactured in many scenarios.

There will be some people in group 3, to be sure. In fact, a lot of owners might be technically in group 3, but they will set a high price. If the going rate for a basic car is 50 cents/mile, they might set use of their car at $1/mile or $2/mile or even more. That way they will only be called upon when demand is very high and the spot price surges up. If it surges up gradually, they will also get a warning so they can take their car off the market (or change its price.)

Unlike Getaround, if somebody does use your car, and even takes it on a long trip, you will always have the option of hiring another car if you change your mind and need a ride. Of course, this other car will also be expensive, so you will neither make nor spend the surge rate, so that sounds pretty good.

You will have to keep your car clean and in top condition at all times. At the same time, as prices go up, you can send it to a cleaning depot to get cleaned and ready for service. (Those depots will probably have high demand too.) The depot can even take your personal items and put them in the lockbox you probably have in such a car, or it could put them in a lockbox at the depot, to be restored to your car when it stops being available. So they can make it pretty painless, at least if you don't keep a lot of stuff in your car.

And yet, AirBNB...

When AirBNB first began and moved past the air mattress stage, quite often the place you would rent was somebody's personal unit. They put it out for rent while they were away. Today, almost all the units I see are dedicated ones which nobody lives in.

Over time, the trend has been towards dedicated units which are run as a business. They are a more reliable and uniform product, with more dependable economics. (On AirBNB some people actually seek out more quirky properties including real lived in ones for a more authentic experience. Nobody wants that in a car.)

Combined with hotel rooms, it is clear that the vast bulk of room nights are dedicated rooms for hire. Of course, it is harder to prepare a lived-in unit for rental than it is a car.

Using up your car faster

Today, privately owned cars average about 10,000 miles/year in their life, and so they wear out by both the year and by the mile. Taxis wear out by the mile, and are gone in 5 years. If you hire out your robocar, it will accumulate more miles, faster. Fairly soon it moves into taxi territory, destined to wear out only by the mile. Once it reaches that point, it is no longer an idle asset that is being wasted if it sits unused, it is a consumable asset, and you want to extract the most value for miles it travels.

There is a limited interest cost and parking/storage cost, but as I analyzed in this article on robotaxi off peak prices that factor is fairly minor.

Two factors push on the decision to hire out. If you hire out frequently, which is to say more than about 25 miles/day, you're really in the taxi business but have a car you can use for your own purposes at wholesale prices when it's not in use. If you hire out infrequently, you're exploiting your asset, but must go through the burdens of keeping it clean and ready to roll for just a small number of engagements. For example, sending the car to a depot to have it cleaned out and your personal "stuff" put into storage will cost several dollars, demanding that it travel at least 50 miles before you use it like your own personal space again.

Some people might like using up their car faster, which is to say, getting a new car twice as often. (As opposed to the more expensive way of doing that, involving selling a used car and buying a new one.) It must be noted that your new car stays new for less time, with parts of it, including the interior condition, wearing out much faster than if you were driving it alone.

The key thing to understand is that once your view your car as depreciating by the mile, it is not found money to hire it out. Hiring out has a cost (in car wear) and in inconvenience for you (your car is not available to you for a period.) The going market rate used by fleet operators will be their most efficient rate, you will need more than that to make it worth it.

Perhaps the most attractive aspect of hiring out, if it can be done at good rates, is to purchase a more expensive car than you normally would, affording that from the profits on hiring it out. It's still more expensive for your own miles, but some of the features that don't really wear out get the cost shared with the business half of the car.

Of course, taxes get mildly more complicated if you do this.

Being a peak demand car

What may make the most sense is hiring out your car during periods of peak demand when prices go up. Depending on how many cars the professional fleets put into service, this may not happen that often. For a pro fleet, except at times when there literally are more passengers than cars, having a larger fleet is mainly a means to offer shorter wait times, and more efficient fleet utilization. (If your fleet is sized exactly to your passenger load, you will see longer wasted miles getting to customers, and longer waits.)

At some points a fleet may wish to hire an outside car (from a private owner or another fleet) if they don't have a suitable car close to a customer. They might be willing to pay a bit extra to virtually expand their fleet. This will not be their preferred plan, since your car probably costs more than one of their own fleet cars. Your car is a more uncertain product, not maintained or cleaned by them, not fitting their standards, and not wearing their livery. (It might have a digital display on the outside to show their logo, though.)

Even so, some fleets will try to distinguish themselves by promising customers shorter wait times. To keep that promise, the cheapest approach may be to be able to bring in outside cars when needed.

As such, you would be able to tell the system managing your car to set a price. Taxi companies needing a vehicle could look at the prices in the market and see if one makes sense. As demand goes up, they will be more and more willing to pay. A person who wants to rent their car out as much as possible might set a low price. Others might say, "I don't really like hiring my car out, but if it gets to $2/mile, count me in!"

This mechanism will also work during periods of special demand, such as disaster evacuations. Everybody might agree to let their car go to work evacuating people from a disaster (which is not happening in their location, of course.)

Car clubs

The same tools might also enable "car clubs" where a group of car owners get together and agree to allow their vehicles to be used, at the right time, only by other club members. Or a club might contain both member's cars and club-owned cars, which may also be made available to the public when not needed by club members.

Cleaning and special features

One challenge is that a taxi will have a variety of features you would not normally need in a private car. I made note of a lockbox above. A few others include face to face seating or a video conferencing link back to the operations center.

One particular feature would be the tools needed to keep the vehicle clean. I have proposed a taxi will have a camera inside which can see the interior. This camera would have a physical shutter that blocks it from looking at you while riding, but to end the ride, you would need to flip open that shutter. When you did, it would take a photo to quickly determine if you have messed up or damaged the taxi, or more commonly if you just left a bag on the seat. In the latter case, you would collect your item. In the former case, the taxi would go to a cleaning depot and you would get the bill. You can't afford to have the taxi go to a cleaning depot after every ride, so a system of this sort makes sense. This is not a likely feature in a private car, though.

On the other hand, it could be that certain types of "special cars" might exist which people like to own but are not very high demand in taxi operations. This might include special "show cars," high end luxury vehicles that people will sometimes pay a lot to just try out. On the car sharing sites, you will see that some people like to hire out Teslas and convertibles.

Convertibles will be a special case. I think the fleets will readily offer them, but in a less expensive way. Rather than having cars which have retractable tops, they may have cars with hard removable roofs that are removed on clear days, and put back on when rain or other weather is in the forecast, to be used as regular vehicles.

Owned cars may well be more expensive than taxis, and they will probably come with steering wheels, though those will be disabled for most taxi service (though people could pay extra to be able to use them.)

What would you do?

I am curious what readers think. What type of car owner (from the 5 classes above) will you be? If you will be a robocar owner, under what circumstances would you want your car to hire out?


This is a great question, and difficult to answer. I am coming from a different perspective than a traditional commuter. I am a mother primarily. Through the years, my car needs have differed, but primarily kept coming back to the same kind of vehicle: a minivan. A decade ago, my main criteria was that I needed a vehicle in which it was possible to fit an infant car seat and two children's booster seats, as well as two adults. More recently, my main criteria was that I needed to fit 6 basically adult people and a wheelchair. In a few years when robocars are widely available, I will probably no longer have a large number of people to cart around routinely, and perhaps I'll be able to think about cars more like a traditional commuter.

So, if robocars were available today, I would be in either category 2 or 4. Four would depend greatly on the schedules and locations of every member of the family working out so that there aren't routine conflicts (people who need transportation at the same time in very different locations). Two would be more likely. I am in the habit of keeping things in the van (shopping bags and emergency supplies, for example). Also, having spent much of the decade+ taking care of the young and the elderly, I would feel weird if an emergency happened, and I had my car rented out to someone when it was unexpectedly needed. That may be a feeling I will eventually get past, but at the moment, it's very much a consideration.

I like the idea of numbers 3 and 5. I really, really like them. But I do not at this point in my life see them as feasible.

Who live in places like Manhattan or London or others where minivan ownership is almost out of the question?

Since there are a lot of families out there with small children, one can be assured the market will provide for them, if it has a solution. Now I agree, a privately owned minivan with everything you need for the family is going to be a winner, but it actually comes with a lot of cost and issues. We might see some very good solutions in terms of vehicles that can show up configured to the needs of your children, and with a bag of "the usual" supplies inside, even. Though you probably need a duffle with you personal favourite supplies. You can keep the vehicle during the day for a pretty reasonable price, probably less than you pay for parking it today in the places that don't have subsidized free parking. People might find that acceptable.

Well, I just bought a new Tesla. The chances I'll put it into such a system, if it existed today, are pretty low. The excitement of owning a new car and attempts at keeping it pristine are high. Show off factor is high. But is quickly going down, especially as more of them appear on the road.

But let's jump ahead to 2023. Five years from now.

My car will have been paid off. The new car smell long gone (it is already gone, since we've put 19,000 miles on our car since getting it in April, it already has a dent and several scratches). I won't care if more curb rash is added to the rims. If there's a problem due to more miles being added, I won't care, because I'll either have a new car or a robocar network where a new car will be at my door in minutes, if not seconds.

I imagine Elon Musk might send me an email saying "hey, don't you want one of my new pickup trucks or one of the new "Y performance models?" Here's a way you can afford to get a new car, while having your payments greatly reduced: "put your old car into the Tesla Network."

I would most certainly do that.

Now, the question is, would I need the new car at all? I doubt it, other than the prestige of owning your own car and personalizing it (I've tinted my windows, for instance, which greatly improves my enjoyment of my car, especially in increasingly hot weather).

Interesting discussion!

We already know people will choose to rent out their Teslas.

Interesting to me is whether cheaper cars can be fitted with similar software soon after the major companies release theirs.

Would a distributed app, built on Ethereum for example, be able.to replac the fees of aggregators like Uber, Lyft, or the Tesla network? If there is a dApp ride network then private owners will be able to cut out the middlemen costs.

The replies to my facebook thread on this suggest that one of the many reasons blockchain won't apply here is that people want a very close and personal trust relationship with the company they hand their car off to. They want an intermediary very much who will deal with problems caused by the person who rides in the car. Now, you can have blockchain transactions where you add an intermediary, or insurance company etc. but then you've defeated the point of the blockchain -- it's there for when you don't want to trust a central party.

For myself, I will be a type 4 user of robocars, which is only incrementally different from my current status. I will very likely own one or two personal cars, and will be much more likely to rent special purpose cars for vacations or other special purposes. I am not likely to try to rent out my personal cars, but I would be much more likely to loan out a robocar. There are two reasons for this. Loaning a robocar will incur much less liability risk, and robocars can be used by people who cannot drive, making it possible to give a ride to such a person without having to accompany them.

It isn't clear to me that a company providing the infrastructure for robocar rentals will benefit from including privately owned cars in their fleet. A company is likely to prefer maintaining the cars themselves, rather than trying to keep track of the performance of a large group of private owners. Possibly such a company might require owner-renters to contract with them for maintenance services so that the company has control and records for liability reasons.

I would like to hear more about what you imagine life will be like for us category 1 people. Will driving my 1974 VW camper van, a 50 year old piece of living history still be possible. If not, I don't want to live in that world. :(

You will still get to drive. It could be that almost nothing changes. However, I have identified a few possible things, some of which you might not like:

  • Some lanes might get designated as only for robocars on some roads. You will still be OK in the other lanes.
  • It's possible certain special areas are designated for robocars only, or electric robocars only. Like the regulated streets of European old towns which require a permit today.
  • Perhaps most scary, the punishments for tickets or accidents may get stronger. Ie. if you cause a serious accident or get a DUI you may lose your licence much sooner.
  • New types of traffic regulation will arise, applying to all cars, not just robocars. You might need to use a smartphone to navigate in the city, and obey rules designed to meter traffic. That's not caused by robocars but by the growing need to meter traffic.
  • Your insurance might go a little bit lower, just because if traffic becomes more regular the odds of you being in an accident decrease.

I'd probably go with 2 assuming that "strictly for your private use" includes allowing use sporadically by friends and family.

Alternatively, especially if it can be done in a tax efficient way, it might make sense to have the small business I work for own a car (or even two). We somewhat frequently have clients with mobility issues, and being able to send out our robocar to pick them up would be great. Whether or not it'd make sense would depend on price, and also on how it'd be handled tax-wise to allow one or more employees to use the car for commuting or even as their main car.

I should add that this assumes a car that can drive itself with no assistance from a licensed driver. I think that's still several years away, especially for consumer purchase. By that time my kids will probably have drivers licenses and maybe even cars, which might move me to the category of reducing from multiple cars to one. Depends how many years off we are, and who is still living at home.

Add new comment