Autolib imagined. Credit: Paris City Hall, 2008

The concept of selling mobility on demand rather than cars themselves may be finally gaining some traction. Remember the stackable urban rental cars proposed by GM-funded researchers at MIT last fall?

Well, Paris is working hard to make that vision a reality. The French capital is gearing up to offer the auto equivalent of Vélib, a distributed bicycle-rental scheme that provides more than 20,000 bikes at more than 1,400 sites across the city and the suburbs. Paris mayor Bertrand Delanoë announced in June that the city will place 4,000 small electric cars at 700 Autolib pickup points around Paris and the suburbs starting in 2010. And according to business daily Les Echos (story en français), train giant SNCF is vying to operate the Autolib points out of its train stations, which are distributed across and around Paris.

And now, the city may finally have a solution to a potential game-killing problem: the uneven distribution of vehicles as cars pile up at popular destinations. Parisians are well aware of this problem. By midmorning, for example, as Vélib stations at the periphery of the city empty out and those downtown jam up, it's not unusual to see trucks redistributing the bikes to counter the tide. That's easy enough with bicycles but harder to envision with even small electric vehicles.

The city's solution? According to a leaked document reported by auto-news website Caradisiac (again, story en français), the plan is to simply have users declare their destination upon checking out a car. In response, the system will determine the closest Autolib point with a free spot for drop-off and reserve that space. No news on solving another potential problem for Paris's Autolib scheme: the name. Lyon, which beat Paris to the bike-share program with its own vélo'v, already sports a conventional car-share program called Autolib.

Could a similar scheme work in the U.S.? Issues of Forbes magazine that will appear on newsstands next week tout the MIT City Car concept as the embodiment of a new car-sharing direction for troubled automakers. City Car codesigner Bill Mitchell of the MIT Media Lab's Smart Cities group adds to the drumbeat in a recent editorial for architecture website BD. "People don't want cars, they want personal mobility," writes Mitchell. He argues that, rather than bailing out car firms, governments should be radically rethinking urban transport around ultralightweight battery electric vehicles (EVs). To provide mobility most efficiently, says Mitchell, we should

. . . organise urban electric cars in mobility-on-demand systems like the Vélib bicycle system in Paris. Racks of public-use cars would be provided at closely spaced sites across the service area. If you want to go somewhere, you walk to a nearby rack, swipe a card, pick up a car, drive it to a rack near your destination, and drop it off.

GM's new electric car, the Volt. Credit: GM

There's still no official word (it's expected by the end of the year), but it looks as though A123 Systems, a company based on a remarkable new battery chemistry formulated at MIT, won't be supplying the batteries for the first generation of GM's new electric car, the Volt. The contract, according to a couple of news reports released in recent weeks, will go to LG Chem, a Korean company.

GM had considered A123, a startup with no large-scale experience manufacturing automotive batteries, in part because A123 had developed a novel battery chemistry that produced very powerful, safe, and long-lasting batteries. So, why didn't the company get the contract?

First, some background on why A123 was in the running in the first place. A123 replaced the cobalt-oxide-based electrodes of conventional lithium-ion batteries with new nanostructured iron-phosphate electrodes. These phosphates are inherently safer than the cobalt-oxide-based chemistries, which have been known to suddenly burst into flames, destroying laptops and cell phones in the process and leading to massive recalls. The conventional cobalt materials also don't last very long--that's why laptop batteries have to be replaced every couple of years. The capacity of A123's batteries, in contrast, doesn't fade much with use. Safe, long-lasting batteries are essential in cars, where they're expected to survive abusive conditions for a decade or more. Ultimately, cost is the biggest issue, and A123 has advantages in this area as well. Safer materials ultimately reduce costs by decreasing the need for redundant safety systems (such as those used in the Tesla Roadster). What's more, longer-lasting materials reduce the need to oversize the batteries to make up for fading capacity over their lifetime--something else that reduces costs. Finally, by replacing cobalt with iron, A123 also reduced the cost of materials.

LG Chem uses a manganese-oxide-based electrode, which is less inherently safe than A123's phosphate-based materials. The company uses other modifications to the cells, including a novel separator between the positive and negative electrodes, to make up for this.

Here are some guesses about why A123 didn't get the contract (if indeed it didn't).

GM may be betting that LG Chem is more likely to supply packs on time. LG Chem is a bigger and older company than A123, a startup founded in 2005, and it has more manufacturing capacity. What's more, Continental, which packaged hundreds of A123's battery cells into a large battery pack, was late delivering packs to GM for testing. Getting the Volt out on time is a big deal for the cash-strapped automaker, which is counting on the Volt to change its image and help turn around its sales. After disclosing that only one of the two battery companies would get the Volt contract, GM vice chair Bob Lutz has reportedly explained that "we feel that at this point we have a lower risk with the one company."

Chem's battery pack might be cheaper. There are a couple of reasons why the many cost-saving features of A123's batteries may not have led to a lower-cost battery pack. First, while replacing cobalt with iron reduces materials costs, working with nanoscale powders is very difficult and can add to processing costs.

Second, the design of the Volt may not take best advantage of A123's cells. The Volt design calls for far more battery cells than are actually needed to supply the car's 40-mile electric range. The pack has a capacity of 16 kilowatt-hours, or 2.5 miles per kilowatt-hour. In comparison, Tesla Motors is selling an electric car that gets 220 miles on a 53 kilowatt-hour pack, or more than four miles per kilowatt-hour. A direct comparison between the two isn't possible because they use different battery chemistries and have vehicles that don't weigh the same, and because the Volt is designed to operate like a hybrid after the first 40 miles, which requires keeping some battery charge in reserve. But the difference shouldn't be this much. According to one GM engineer, 12 kilowatt-hours should be plenty of energy. The extra four are essentially for insurance against battery degradation, so that at the end of a decade, the Volt still gets 40 miles out of the battery. A123's batteries may not need this kind of insurance, since they are so stable. That stability could make it possible to use fewer batteries than is possible with other chemistries, cutting costs. But GM requires A123 to supply the extra cells anyway. That could be wise, since better tests are needed to guarantee battery lifetimes, but the result is that the potential of A123's innovations isn't being exploited, so the packs are likely more expensive than they need to be.

Not getting the contract, which is reported to be for 50,000 battery packs, can't be good news for A123. But it's not the end for the company. It is still in the running for the next-generation Volt. What's more, the company is working on batteries for 18 other vehicles.

Better Place (formerly Project Better Place), a company that plans to develop electric-car infrastructures for Israel and Denmark, has now announced plans to do the same in a much bigger country: Australia. The plan is to eventually make it unnecessary for Australia to import any oil.

If it succeeds there, the company's model could work in parts of the United States, too, such as the West Coast or the cities from Boston to Washington on the East Coast.

Better Place has proposed ways to overcome the limitations of today's technology for electric vehicles--namely, the cost and recharge times of batteries. To keep down initial costs for customers, the company plans to sell cars in much the way that mobile-phone companies sell phones: with a subsidized low cost and a monthly plan. For the cars, the plan will pay for miles of driving, not minutes of talk time. The company also plans to install networks of charging stations, so that drivers can keep their cars topped off during the day, and battery swap stations along highways, where drivers can exchange a depleted battery for a charged one on long trips.

The plan seemed to make sense for Israel and Denmark, relatively small countries where such networks could be easily installed and where government policies heavily favor electric cars. But Better Place's CEO, Shai Agassi, has said that it could work in the United States as well. (See this video.) Rather than connect the whole country, however, the plan would be to connect certain urban centers, such as those from Boston to Washington, DC, or from Los Angeles to Seattle. Government policies would still be needed to make the plan economical.

In the announcement of the Australia deal, Agassi emphasized that if the system can work in Australia, which has more car ownership per capita than the United States, it could work in the United States.

Not everyone is so optimistic.