» An all-electric, point-to-point system could revolutionize how we think about the automobile and significantly reduce the need for private cars in our cities.
American urbanites have already become quite familiar with the concept of car sharing through the rapid expansion of companies like ZipCar and I-Go; the ability to rent a car at a reasonable price at any time from a location within walking distance of home or work has dramatically reduced the need for at least some people to own private vehicles, since it covers the gap in service not provided by transit: Trips that are out-of-the-way, that require moving heavy goods, or that occur at inconvenient times. This is great for cities and for people, since not only does it reduce the need for parking, but it reduces vehicle capital expenses for everyone, since the cost of purchasing the car is effectively shared among many households, not just one.
But car sharing has always been limited in one significant way: It forces users to return cars to the place where they found them. This makes it impossible to use car sharing for one-way trips and makes it difficult to use a shared car for trips that last more than a few hours — because it simply becomes too expensive. And the environmental credentials of existing systems are limited: They mostly rely on the same old gas and diesel-powered cars regular automobile users own.
In several cities around the world, however, a new form of car sharing is being developed that offers users both one-way trips and electric vehicle fleets. Together, these advances could increase the number of people choosing to abandon their private vehicles in favor of shared cars.
Paris took the most significant — and most visible — step forward this week with the introduction of its new Autolib’ service, modeled after its highly popular Velib’ bike sharing system. A fleet of dedicated electric cars (called “Bluecars”), manufactured by French company Bolloré and featuring two doors, four seats, and a GPS system, have been distributed across the city; by June next year, there will be 1,740 such vehicles spread across Paris and 45 suburban cities and towns (eventually, 3,000 vehicles are planned). Cars will have the ability to travel up to 250 km on one four-hour charge.
Unlike most other car sharing networks, but like Velib’, Autolib’ vehicles will have to be parked in dedicated parking spaces that feature electric plug-ins. In other words, while the system will allow travel from one point to another, at each end people will have to find a specific station where charging is possible, since users will be required to plug the cars back in once they arrive. There will be 1,120 stations by next year (up from 33 currently), with 508 planned in the city itself. Some of these locations will feature distinctive booths (such as the one pictured above) that allow people to scan in their IDs, driver licenses, and credit cards in order to get started using the network.
Users will be charged higher annual fees (about €140/year) than do services like ZipCar, but will be compensated with relatively lower hourly fees, which will run about €10/hour.
The City of Paris has contributed the funds to install these stations, with Bolloré covering the costs of the cars. The service is expected to be operationally profitable over the medium-term; risks are being covered by Bolloré.
Paris is not the first to pioneer point-to-point car share at a large scale, though it is the first to do so with a fully electric fleet.* Car2Go, a subsidiary of car manufacturer Daimler, has since 2008 provided such a car sharing option first in Ulm and then in Austin, Hamburg, and now Vancouver with two-seater Smart Cars, which Daimler makes. The expansion of Car2Go suggests that the point-to-point model is economically profitable and that people appreciate the concept.
Unlike in Paris, where expensive installation of electric charging devices and station street furniture has been part of the process, Car2Go allows users to leave cars in any parking spot in the city, not a problem since users can simply fill up the cars at gas stations.**
Car2Go, however, plans a move into electric vehicles, beginning with a 300-vehicle fleet in San Diego at the end of this year. Though Car2Go will not be installing charging stations there itself, another company, Blink EV, will be plugging 1,000 stations in for any users with electric vehicles (Autolib’ will allow people with private electric cars to charge in Autolib’ spaces as well, as long as they’re signed up for the service). Customers using Car2Go there will not be required to plug the cars in (they will have to be charged every two days or so), which leads to questions about who will do so. More maintenance staff to do such work will ultimately mean higher prices for the customer.
For both systems, important questions need to be answered: How can we prevent people from using these networks for one-way trips to work, instead of the public transportation system? Will cars become physically concentrated in some areas of the city and have to be redistributed, as are the bikes in most bike sharing schemes?
The pricing system in Paris will likely disuade people from using the cars to get to work — unlike with bike sharing, which generally has no fee for members in the first half hour of use, Autolib’ requires payment for any use. Thus cheaper transit will remain the most appealing option for people in cities with good rail and bus networks. In less well-served markets, like Austin, one can imagine more people who can afford it using this type of car share for daily commutes.
What is clear is that the networks will have to be constantly monitored to ensure that everyone has easy access to vehicles at any location and at any time. After all, the primary users are people who would have taken their own cars or hailed a taxi otherwise. People are paying significant sums to become members (certainly more than they are for most bike share programs), so they need to be able to rely on the network.
One thing these systems could do is make way for more public space. Paris expects each Autolib’ vehicle to effectively substitute for five private cars. With 3,000 vehicles planned to be included, these would in theory substitute for 15,000 private cars — meaning that if each car accounts for about 1.5 parking spaces, the region could eliminate 18,000 spaces ((15,000-3,000)*1.5). If parking spaces are on average about 180 square feet, that’s a not-insignificant 3.24 million square feet of urban space that can be reused for something else, like parks or in some cases new buildings. The €56 million public subsidy to install the stations (at €50,000 each) seems a small price to pay to free up all that room.
* Other providers, such as Nice’s Auto Bleue, have offered all-electric fleets, but they require users to return vehicles to the places where they picked them up.
** This difference is similar to the contrast between Paris’ Velib’ and Montréal’s Bixi bike sharing networks. Whereas Paris invested in stations that required tearing up the street, Montréal’s system relies on a series of pods that are set down impermanently on basically any flat surface and which can be later moved. The latter approach is cheaper, less physically intrusive, and more adaptable. On the other hand, the physical presence of the built stations of Autolib’ signify a permanence and easy-to-find nature that cannot be replicated by park-where-you-want systems like Car2Go.
Image above: An Autolib’ station under construction in Paris, from City of Paris
53 replies on “Car Sharing 2.0 Leaps Forward in Paris”
Cool if it works…but color me skeptical. This is a logistical nightmare.
When there are supply imbalances with bike sharing, there are ways to fix that: tow surplus bikes back to where the supply is low. Not extremely efficient, but it works. But it would take either an army of tow trucks, or parking garages on every corner of the city to keep this kind of supply balanced.
I don’t buy the “it wont be so bad because it will price out commuters” argument either. For one, price is relative. If you work in downtown manhattan and you end up working late or overtime, many companies will pay for car service…certainly not a typical commute option, but there is enough volume that most car services are running empty back to downtown manhattan in the evening hours.
But even if nobody ever commuted with the cars, they still typically are more active and away from their homes during the day and they all wanna go back to their homes at night. That is still a cyclical imbalance.
I’ll be watching this one closely.
The question of whether those cars are useful for commuters depends on the comparative costs of the car share and buying a car. For example, driving 45 minutes one-way ten times a week forty-seven weeks a year costs about €3,500 a year, which is about the same as the cost of owning a car. For that money, you might as well buy a car and have it available not just for your work trips.
If I had a license, those cars would be most useful for me for moving, or other trips involving moving many items – in other words, what ZipCar currently does, but without needing to return the car to the same place I rented it from.
Exactly, I don’t see a big role for these services.
A one way car sharing system is useful in the same way that one way bike sharing is useful.
Say a trip to Costco or the Home Depot. You dont need a car to get there….but you’ll probably want one on the way back. With zipcar, you pay for both direction + the time you’re in the parking lot, and nobody else can use the vehicle.
So who drives the car TO the Costco?
The platoons of hired drivers that take mass transit to the outlying stations and drive the cars back. Shouldn’t be all that more expensive than taxicabs and you get the sheer joy and pleasure of chauffeuring yourself through city traffic
One way car sharing already exists – it’s called a TAXI. It has labor costs (the driver), but no parking costs. And no danger of people damaging the car (accidents or otherwise) and then returning it without taking responsibility. Bike sharing works only because the parking and damage costs are much smaller.
Eric, the labor costs of a taxi are simply unsustainable for many circumstances a car would be ideal.
For instance: hiring a cab to drive you whenever you want to a nearest hiking departure point, staying there idle for anytime between 3-6 hours, then driving you back. Or going in a Saturday errand through Home Depot and Ikea.
In developed countries, the manpower cost of a car fleet is far higher than the vehicular ownership and operation cost.
Taxis don’t work that way. They drop you off at your destination and leave. They then go and service other fares. When you need a taxi again, one whisks you away. You pay for the time you are actually using one.
In Singapore, the daily charge for a taxi is about S$100 or so. (When I lived there it was $80-90, but it periodically goes up.) The income of an average taxi driver is lower than that. It’s not an order of magnitude lower or anything, so eliminating the driver can reduce a noticeable chunk of the cost, but it’s still significantly lower.
But getting a taxi home from shopping is fairly common in the UK. Many major supermarkets have taxi courtesy phones near the exit, or adverts from local minicab firms. For those reliant on public transport a bus to the supermarket and then a cab back with the weekly shop is a regular occurrence.
This has declined somewhat in recent years in the UK because of the big take up of internet home delivery now available from every UK chain. In fact home delivery has proliferated across all retail chains, not just supermarkets, in the UK. Many large department stores are also looking a mini stores in smaller towns which will primarily act as ‘click and collect’ sites for people who are not in for home delivery.
There’s a good comment thread on GGW on this sort of car sharing because Car2Go (Daimler’s one way car sharing using Smart forTwos) is planning to come to Washington.
One of the more interesting comments was from someone in Vancouver, which already has Car2Go, that he was a worse driver using Car2Go since he was aware the meter was running. He’d run a yellow light that in his own car he’d stop for, for example. There are always unexpected side effects.
network supply imbalances of massive deployment of shared cars can be tackled relatively easy with reverse pricing.
If there are cars needing to be driven at a certain time of weekdays from point B to point C, give deep discounts on that pattern, or even revert payments (= give credits) for members willing to drive cars on that route.
This idea wouldn’t work with bikes because they are not traceable, and usually don’t cost substantial money to use on a single run to make reward “passive drivers” feasible.
As for the floor space, I disagree with some calculations. Underground parking space doesn’t have much use besides parking garages anyway. For mechanical reasons, machinery on high-rises is usually placed on the top, not lower underground floors.
I think large-scale car-sharing (and, further down the road, Personal Rapid Transit) schemes have huge potential.
Once a critical mass of vehicles is deployed so that everyone is close to one of them, it makes the system trustworthy for what cars are unbeatable: off main pattern, uncommon, late night or high-comfortable journeys.
Increasing the rotation of cars, liquid fuel or eletric-based, can also bring a shift towards efficiency, more or less like the concepts used to design an industrial clothes washing machine running 14h a day are different than those applicable to designing a home appliance level front load washing machine.
A second benefit of a network with high use of car shares is that it allows many more hybrid trips with use of cars and trains.
For instance, a metro-wide scheme with sufficient density might well solve last-mile problems that doom so many transit problems.
If there is a massive park-and-ride station with frequent (5 min or less) fast shuttle trains downtown, a robust car sharing scheme might allow for dozen of thousands of people living a wide catchment area not to drive downtown. Then, suddenly you don’t have the need to drag the feet of high-capacity systems with low-performance (money and time) local buses stopping everywhere in bizarre routes. For that, there is car share.
Excellent combination indeed.
On the other hand, a massive park-and-ride guarantees the system will only be used by peak-hour commuters because the stations will have no commercial development and be unwalkable.
What’s happened in Japan is that as buses became intolerably slow, people substituted bikes as the main means of getting to the train station. Cars are not a possibility, since the railroads have much higher-value uses for the station areas.
There is no reason to apply Asian, particularly Japanese, space constraints on transportation planning for Europe, let alone North America.
The problem you mentioned can be solved with dual-station model, in which each destination has one main station with develop around it and just a bit down the line a massive park+ride (and nothing else) sibling. If frequency is high enough, final destination of each train can be alternated.
Bicycles can be useful for certain circumstances, but the % of trips, considering combination of weather x hauled luggage/cargo x body ability, that can be taken by bike is significantly lower than those that can be taken by car.
Okay, so you’ve just doubled the number of stations, making transit uncompetitive on time, and helping ensure frequency is not in fact going to be high enough.
Transportation planning in Europe deprecates park-and-rides just as much as in Japan. The emerging practice in Germany, Switzerland, France, etc., is to locate stations in walkable areas and small-town downtowns and have timed connecting buses. It’s the opposite of what the MBTA does, which is deliberately locate stations away from the cities the lines pass through and build park-and-rides for one-way service.
Nearly all rail stations in Europe have been located in dense urban walkable neighbourhoods. They were built before the car. Park and ride varies in size in Europe, and rarely like as big as seen in the New World. But it does exist.
Many UK stations have car parks (mainly built on small slips of land that used to be the Stations goods(freight) station. Many prefer to drive to an express station, rather then use their local station. Railheading, as it is termed, exists because, while a local stations half hourly service may be fine if you live in walking distance, your risk of missing a connection increases the further you live from that station. If you have to drive, or rely on a bus, you might as well go to the main station where you can catch an express to the main city, or you have several different services to choose from, so it does not matter if you miss one train, another will be along in 10 minutes.
Where a town can not accommodate, car commuters in it’s historic core, often a park and ride station will develop to catch commuters from surrounding towns and villages too small for rail service or a good bus service. In some places where this is on a main line, rather than on a branch line, the park and ride site has ended up with more express services than the historic core.
I suppose if all you are running is a few commuter trains a day then such choices have to be made, but when it’s a complex mix 10 to 14 services and hour a along a main line then a complex pattern of services can be used to access different markets.
Car2Go in Austin has completely proven itself. I have used it as a secondary vehicle and, after some initial glitches, have found it a better alternative to a rental car or public transportation. I was so impressed, I even blogged about the experience (http://blog.heimsath.com/blog-0/?Tag=Transportation).
The service (and parking area) is limited to the downtown core and immediate surrounding neighborhoods, so there isn’t a problem with the feared commuter log-jam. A look at the map at any given time shows that the distribution is rather disperse. There has only been one time where I did not find a vehicle in walking distance.
Car sharing, as it already exists in its traditional form, has revealed itself most likely to displace a 2nd car ownership in the household.
I wouldn’t worry about commuter use. The existing Car2Gos define a service area, within which cars must be left (you can drive the cars outside the service area, but have to come back inside to drop them off). That means that the cars aren’t physically available to suburban commuters. The assumed service area in Washington is likely to be the District or the ten mile square (which includes Arlington) or at most some bulges out from that to incorporate e.g. Bethesda. In Paris, the service area will almost certainly be the city, the twenty arrondisements. If there were a one way car sharing service in New York, the service area would probably be Fortress Manhattan. The point about these service areas is that they’re thickly serviced with transit already, particularly Paris. The cost of car sharing is going to be considerably above the cost of transit. One might use it occasionally, but most of the time you’re going to get to work by transit, as you do now and always have done.
The trips which will be diverted to one-way car sharing will be those currently serviced by taxi. Car sharing tariffs are way, way below taxi tariffs. Shared cars are likely to be at least as easily found as taxis. And you won’t have to tip a surly driver. If Car2Go did come to Manhattan, look for the value of taxi medallions to fall precipitously.
That is an interesting observations. I like the idea of getting rid of most taxi drivers when passengers can drive their cars.
Like train drivers, the less manpower employed in numb tasks in transportation, the better.
Enhanced navigational devices that are smart can easily eat up the other advantage of taxis: knowing “the secret” about jams, best routes, addresses, complex trips.
Agreed with @Alon and @Eric that one-way sharing seems like a novelty for most markets, with inefficiencies vastly outweighing the various benefits. Given a city with taxis, two-way car sharing, and a decent bike network, I can’t imagine these services finding a niche.
The real shake-up here will finally occur when self-driven cars make their way to market…
Wait, where am I saying that the inefficiencies outweigh the benefits? Are you conflating me with Danny?
Personally I think those cars are pretty much useless for commuters, but have benefits for occasional non-peak trips involving heavy luggage. They’re like a taxi service that you have to drive yourself. I don’t think anyone can say with a straight face that taxis are causing people to drive more.
My point is that they are just like taxis… except without much of taxis benefits. They’re only marginally cheaper than a taxi and require that you A) are close enough to one initially and B) can find parking close enough to your destination. Your moving heavy objects without needing to return the car is an extremely limited use.
I think many commenters here, not you Alon, are seriously underestimating the logistical problems of reserving these and maintaining a sufficiently-distributed network. Sure $12/hour sounds great until you factor in spotty pickup and drop off. Pricing routes differently will make for a very slowly correcting system.
Off-peak time, at night, in the countryside, one hour get you very far away.
Remember in many cities taxis can’t be hailed from streets and waiting times on busy nights can be as high as 30 minutes from the placement of your call.
Right, except that one-way rentals would not work for the scenario you propose: the car would then be in the middle of the country, at night. So either rates would have to raise considerably, or leaving cars there would not be allowed.
Marcella is right about off-peak and at night. Even within the city, I can drive from Georgetown to American University in less than ten minutes (along Foxhall Rd.) non-rush hour. That’s under $4. For two people that’s even cheaper than Metrorail and I wouldn’t have to walk to/from Metro or transfer within Metro (which at off-peak headways is a pain). A cab, by the way, will run me something over $10 depending on the tip.
Jim, you’d have to pay for 1/2 hour minimum for that ride (there’s no way they’d do 15 minute billing). Over that $10 taxi, You’d really rather pay $7 for a car that may or may not be there and then park at a designated spot that has a chance of being full?
Here are the Car2Go rates for Austin. I do not see any mention of a minimum charge. I see mentions of maxima.
Oh and just to confirm: Your credit card will be billed by the exact number of minutes you use the car.
And from the same page:
Several people have assumed that the car sharing company would not make arrangements with the city for there to be dedicated parking spaces that they can use. As you can see, this, too, is a false assumption.
Well, I’ll be. I stand corrected.
There are still major issues with this model, especially w/r/t parking at your destination.
Does Car2Go pay some sort of “park” anywhere fee to the city?
Assuming they didn’t make arrangements in private parking facilities (one wonders whether the city would like revenue from, say metered or free street spaces, as this is a commercial venture – in Chicago, for instance, commercial vehicles cannot park over night on residential streets).
I’m more and more intrigued by this. Seems it would work best somewhere walkable with good transit but an occasional need for a car. I can’t see it for commuting in an area with heavy traffic congestion, or, in fact being useful in a region with constant congestion. There are people who abhor cabs, for various reasons, and in cities with limited taxi service.
I’d be interested in hearing more from a user – since it’s spreading there must be some utility to it.
In Washington, apparently there’s a brouhaha that the city will auction off parking spaces for carshare instead of giving ZipCar a monopoly.
The city has auctioned off the carshare parking spaces that ZipCar used to get for free. There were 84 of them. Zipcar ended up with 14. Hertz On-Demand and Car2Go bought the rest. We don’t know how many each got.
David G., if you take really long trips with a cab, outside “city limits”, sometimes the fares go higher, much higher, to account for the time it would take for the driver to go back to his catchment area.
If you are going with a cable to a middle of nowhere destination, the driver will want to negotiate the fare.
Try getting a cab in Washington for a day-long trip at Shenandoah National Park, leaving 5am and returning 4pm, and see if (s)he will charge you a regular DC-based fare… never.
If you take really long trips outside of the city there are many companies willing to provide you with long-ish term rentals of automobiles. Hertz, Avis, Enterprise….
But what about people who can’t or don’t want to drive, and don’t want to take, say, a bus tour?
The real story here is that electric cars are finding their way to the uses where they make sense – high usage vehicles rather than lightly used private cars.
Electric taxis, delivery vehicles, and shared cars make sense. The one-way variation is only that, a new variation on car sharing that we will learn more about with experience.
One shared electric car can replace 5 or 10 gasoline powered cars. That is the kind of transformative change needed given the end of cheap oil and the climate crisis.
This economic analysis is correct.
However, it has been stated that to be a viable taxi for London, an electric car will need more than a 100 mile range. Tesla Model S (300 mile range) for taxi services? Actually, it makes a lot of sense.
The car must generally have a range sufficient to cover the day’s usage, and with a high-usage car, this is an issue.
This is an excellent development. I use Zipcar sometimes and the number one problem with it is the two-way nature of the service. I am constantly running into times I would like to take a one-way trip, or times I want to make a long round-trip but don’t want to pay for the hours I spend at my destination.
I would also welcome the opportunity to pay a higher annual fee in exchange for a lower hourly rate. Right now the high hourly fee is a huge disincentive to use and probably leads many people to hold onto their cars. After all, most costs of owning a car are sunk costs, so people rarely consider the cost of each trip.
Anyone who says taxis adequately fills this need must live in one of the few cities where it is easy to quickly get a cab, and/or one of the few cities where taxi fares are not enormously distorted by anti-competitive public policy. This reminds me of Oregon’s law requiring full-service gas–it generates employment, but the attendants are doing a job very few people really demand. Similarly, given the choice to drive your own cab and pay a lower price, most people would take that chance. There are exceptions, such as when you are drunk or don’t know the area, but most of the time carshare would be preferable.
Lastly, I’m so happy people are realizing that the great potential of electric vehicles is in precisely these kinds of short-distance applications. This could really help to encourage more affordable developments in urban areas that do not require owning a car. Things like this also justify public expenditure on electric charging infrastructure, since it has a broader benefit, rather than just helping out wealthy people who can afford new electric cars.
or times I want to make a long round-trip but don’t want to pay for the hours I spend at my destination.
There are businesses that meet the need for day long, week long etc. rentals. Hertz, Avis, National, Enterprise, Dollar….
zefwagner, in Europe, where the marginal costs of mobility (namely fuel) are much higher, you don’t find ZipCar-like plans.
They will usually charge a low hourly rate + a distance-based fee. It makes it more attractive for errands, less attractive for long one-shot trips, where renting from Hertz or Europcar might give you a better deal.
I think this is a very different thing from ZipCar. The use of the term “carsharing” is misleading.
With ZipCar, I have a need for a fraction of a car. So I and a group of neighbours share a group of cars. We really only share the cars that we can easily access. ZipCar makes its money by ensuring that the number of cars available to us satisfy the sum of our fractional car requirements. Those cars are the ones that actually substitute for a personal vehicle. The ZipCars in, say, Northeast Washington are irrelevant to me. I don’t care if they exist. I’ll never find myself driving one of them.
Car2Go or AutoLib on the other hand are more like PRT (Car2Go’s forTwos even look like pods). They aren’t automated, there is no fixed guideway and I can’t go to a station and call one up at the touch of a button. But all 300 Car2Go’s cars in DC will be equivalent, they’ll all be equally available to me. I’ll take whichever is close to me when the urge to take one strikes me. I’m no more sharing the cars than I’m sharing the city buses or subway trains. The Car2Go system provides a publicly available transit alternative which will take just me from where I am to where I want to be (as long as both are within the system). Which sounds a lot like PRT.
You cannot park a CAR2GO “anywhere”.
In downtown Vancouver there are some designated spots reserved for CAR2GO. You may also use designated residents parking spots on city streets in the operating area (roughly the northern half of the City of Vancouver). Finally, there are are places where any car can be parked legally on the street, and those are also available (of course) for CAR2GO. While you can leave the car, and take the key with you, as long as you want to keep paying rental charges by the minute, to end the rental the choices are more restricted. More details can be found at the car2go web site.
On the other hand, just the other day, I spotted a Swiss Mobility car in Avignon (at least 400 km from the Swiss border). OTOH, it could have been someone buying a phased-out Mobility car, and the markings were not removed…
Just to the discussion: The partnership between public transit agencies and car share operators can be very successful, particularly when for example annual pass holders get a discount on the car share rates. In fact, I know of several families in Switzerland who cover all their transportation needs with annual passes for public transportation plus membership at Mobility car sharing (plus occasional taxi rides).
@Max Wyss: I lived in Switzerland with my boyfriend for a time. We did have a car and a mobility subscription.
Mobility CH has some plans that allow you to use their cars as normal rental units, particularly over weekends, holidays, summer when demand is lower. It’s a like a way for them to lock-in the costumer and avoid him/her to go shopping on Hertz, Sixt, Europcar etc.
I think car sharing is a novel idea, however they should be private businesses. In my opinion, government running any sort of business is always bad news.
-Chris, Realistic Flight Simulator Reviews
I agree with Chris’s (not me) post above; there are enough venture capitalists itching to fund a well-construed business plan such as for shared electric cars that getting private-sector funding shouldn’t be a problem.
Re: how to prevent these from being used for commuting? Congestion pricing could, but as long as the cars are all-electric, and would otherwise be sitting there, and if public transit systems are overloaded (such as on the #4 train line in Manhattan), what’s the concern?
In this case, it should be no problem to have this be done by private businesses. It’s not a natural monopoly.
However, government can run businesses excellently; the Chris above is simply being doctrinaire and stupid. The businesses best run by government are natural monopolies like public water supply (yes, it’s a business).
I’m still going to drive my personal automobile in and to the city in that I really not nuts about being a publicly owned car that a million strange people have been in.
If you don’t own a car and live in a city this is though a great idea.
Just to add to the skepticism about the one-way system:
The service will surely be attractive to (some) customers, but I agree with the operational issue mentioned (redistribution and availability), which compromises profitability – and the issue of finding a niche and competing with PT and taxi
With a couple of months hindsight, here are some extra issues that have become clear
1) one-way cannot replace the private car, since you can only drive within a restricted urban area (Autolib Paris en Car2Go Amsterdam), exactly to keep recharging and redistribution manageable for the operators – So in effect, you are encouraged to drive short trips in the city by car, which would preferably be done by public transport, bike or walking
2) it may replace cars, but it might actually encoruage car use for the intraurban trips – see above
3) the claims that each car replace 5 cars is simply copied from traditional two-way carsharing and not substantiated for one-way;
4) the space-saving claim is equally contentious. Paris autolib has plans for 3000 cars and 6000 parking spaces, for the operational reasons mentioned above. So we can start by halving the replacement rate. If
Oh, and I forgot:
5) carsharing with electric vehicles (two-way or one-way) are usually heavily sponsored by makers of electric vehicles or charging stations. As a recent EU-study concluded (MOMO project): electric vehicles need carsharing more than carsharing needs electric vehicles. Autolib is basically a showcase for French EV-maker bolloré, who is now heavily investing in having its vehicles visible all over Paris and installing charging stations everywhere: even is the one-way-service turns out not to be commercially viable, the infrastructure will still be there.
This is fantastic of Paris to embrace the idea of Car Sharing. It really does benefit those who cimply can’t afford to run a car.