This 15-foot-wide, solar-panel-adorned disc is a floating water purifier, to be plopped into the canals in the Japanese city of Osaka, and into the moat around the city's centerpiece castle. Designed by NTT, these "floating UFOs" can filter about 2400 gallons of water in the 6 hours a day they'll be operational, all the while spewing the newly cleaned and oxygenated water out of a little spout in its back, presumably because these things have the potential to be kind of unsettling, and everything with a blowhole is automatically charming. It's true! Check!

At night they'll just sort of float along, creepily. They've got batteries for when the sun hides out, but only to power a rack of glowing LED lights. [Asahi via Pink Tentacle—Second image from Mainichi Daily]

A project including the joint efforts of NTT, KDDI and SoftBank, Japan's big three wireless carriers, data lines were actually run aside the entire track and use periodic wireless transmitters to connect to passengers.

From inside the train, users should enjoy speeds of 2Mbps even through tunnels, while those waiting at station concourses will be blessed with speeds up to 54Mbps (which happens to be the peak speed of wireless n routers, or about 6 times the speed of the only DSL service to offer wiring in my apartment building). [CrunchGear and Getty]

Ed note: Technically this shot is from China, but shhh, I just liked it.

Japan's NTT Communications is testing a new billboard setup in January that has a built-in pair of cameras hooked up to image detection software that determines how many people are in front of the ad, and just how many are looking at it. It doesn't try to identify individuals, or tailor the ad to specific demographics, unlike some proposed systems.

The way NTT's system works is that it compares the image of passersby to an "average Japanese face" and determines whether or not they're peepingthe ad:

"We gathered together many faces and came up with an average Japanese face, and by using pattern matching the system recognizes faces from the image."

So, uh, does that mean it won't work on white people? [Good Gear Guide via /.]

The data is passed on to other devices as you touch them, even with your clothes or shoes. So you can open an office door keylessly, unlock and start your car, or any of a million other applications currently using swipe-card entry.

It's clearly more convenient than having to fish out a conventional card, and more secure than a wireless device whose signal could be snooped on. NTT even foresees medical applications in the future, since the system could easily and discretely transmit health-sensor data to doctors and nurses as they touch you during exams.

No prices are announced yet, but it will be "a bit pricier" than existing systems. [Sydney Morning Herald]

No, it's not what we had in mind either. The sign's smells will change throughout the day, emitting lemon in the afternoon and "woody" aromas at night. The bottles of chemicals are used in various recipes to create the smells, which are ultrasonically sprayed across a 5,400-square-foot area. The sign can gather updated scents through an auto web download. And for the ambitious hacker, we're guessing some nasty pranks could be programmed for unintentional public consumption. [press release via pinktentacle]

It seems that NTT, the biggest fiber provider in Japan, has a lot of the same issues Verizon does here: they need to get permission from landlords to hook buildings up with fiber, and once they do they still need to convince people to sign up for it. And after all the big apartment and condo buildings are hooked up, they're stuck doing individual houses.

Overall, setting up a fiber optic network is a very expensive prospect with no real guarantees to making all that money back. Without a lot of current applications that utilize such crazy speeds, there just isn't a big enough demand to justify the expense. So why does Japan throw caution to the wind and spend all sorts of cash to set up this speedy network even though it might not be the most fiscally responsible thing to do in the world?

Well, one of the big reasons is that the Japanese government provides tax incentives for companies to do so, while our government has done basically nothing to encourage a nationwide fiber network. Despite the fact that setting up a completely new wired infrastructure is an incredibly expensive undertaking (just ask Verizon), companies in America are supposed to do it all themselves and make it profitable, something that might not make sense with this situation.

Matteo Bortesi, a technology consultant at Accenture in Tokyo, said that "the Japanese think long-term. If they think they will benefit in 100 years, they will invest for their grandkids. There's a bit of national pride we don't see in the West." I'm pretty sure there's more to it than just national pride and wanting to be first, but there's certainly something to be said for setting up a network now that we'll need in a few years. Chances are, eventually we'll have a nice, fast, cheap, nationwide fiber network that will allow us to download porn faster than we could ever have imagined before, but at the rate we're going, it's going to be a while. [New York Times]

Just the scenario of turning off your radiator, heater, TV or A/C if you forgot to when you left the house should be enough. But how about turning on a rice maker when you're about to leave work to go home? Or turning up the heat in preparation for your arrival? You're sure to come up with even cooler applications yourselves. – Jason Chen

NTT's cellphone-operated remote control home system [Pinktentacle]

Smell my web [Akihabara]

Hey, Your Wallet's Ringing

Device convergence in the mobile world typically means something like shoehorning an MP3 player into a handset (sometimes with hilarious consequences). But plenty of companies have have designs on cramming your wallet into your cellphone, too.

Wireless payment systems aren't all that new. The most well-known example in the US is ExxonMobil's Speedpass, which uses a little plastic cylinder on a keychain — though it's also been built into some Timex watches. The cylinder holds a tiny radio transceiver, which a user waves in front of a gas pump or reader at a cash register; and the system then bills the purchase to an associated credit card. The system is simple and easy to use, but its biggest drawback is that it can only be used at ExxonMobil gas stations. If you pull into an Amoco, you re stuck with whipping out the old-fashioned Visa or Amex.

Speedpass, then, acts as little more than a credit card replacement at one chain of gas stations. It may be a marginally useful product, but its limitations illustrate two characteristics mobile payment systems must have to really succeed: high utility and wide usability.

In other words, they ve got to do more than just replace your credit card, and they should be accepted by more than just a handful of merchants.They must be useful, both as payment replacements, but with other applications, and they must be accepted at a wide number of places. That's part of the problem with Speedpass and other similar programs. While it's helpful for ExxonMobil customers, imagine carrying a separate key fob for each brand of gas station, convenience store and supermarket—your keys would be so cumbersome, they d require their own backpack. (Of course, some retailers view the payment systems as lock-in mechanisms, an attempt to get people to shop at their stores exclusively.)

Mobile payment systems are proliferating, but it's still the early days, and (in the US, at least) a viable one is yet to emerge. One attempt, MobileLime, launched this summer, taking a slightly low-tech approach: when users go to pay at a merchant that accepts it (currently about 45 places in Boston), the user calls an 1-800 number and enter the business ID number and a PIN. They then give the merchant the last 4 digits of their phone number which is entered into the system along with the price; the money is then deducted from a prepaid account, or billed to a credit card, which then deducts it from a prepaid account or bills it to a credit card. Not particularly convenient when compared to something you just wave in front of a reader that handles the rest.

All the major credit card companies are working on contactless payment systems, and they're looking at integrating them into phones as well. One Nokia-Mastercard trial put RFID chips into handset faceplates — basically putting the SpeedPass in the phone.

But these trials still don t deliver what s necessary for a truly successful mobile payment system. Credit cards aren't always the ideal mode of payment, particularly for very small transactions —exactly the kinds of transactions where mobile payments are most useful (think buying the morning paper, or a can of soda or pack of gum). But more importantly, credit cards generally can't be adapted for applications that would make them even more useful, like doubling as subway passes, or access control equipment.

Not surprisingly, we have to look to the East to see a better way. The FeliCa system, developed by Sony and first introduced in phones by NTT DoCoMo, has emerged as the country's mobile-phone payment standard; it s also been rolled out on handsets from the country's two other carriers, KDDI and Vodafone. FeliCa's base functionality is as a payment replacement system, where users fill an account with funds then debit purchases, or connect it to various brands of credit cards. Japan s mobile carriers also helped spread FeliCa s acceptance by smaller vendors by setting up a fund to help pay for point-of-sale equipment.

But what makes FeliCa really interesting is that it's not an application; it's a platform that lets other applications run on top of it. This means that companies and developers can write their own FeliCa functions that can be downloaded to compatible phones, constantly expanding its functionality. One of the most popular applications has been to make FeliCa phones work with the Suica contactless IC ticket system of the East Japan Railway Company around Tokyo, letting people use their phones as train passes and tickets.

One application that highlights FeliCa's usefulness as a platform is the Kesaka system. At its most basic level, a FeliCa phone can be used as a key to open a Kesaka lock. But the application on the phone lets the user remotely check if the lock is open, notifies users when the lock is unlocked, and lets them distribute "duplicate keys" to other people's phones. Those dupes can then be set to expire after a certain time, as an extra security precaution.

FeliCa phones can also serve as plane tickets, loyalty cards at stores, movie tickets, and, of course, change for Coke machines. All of this is possible because of the platform approach — nearly any type of transaction or situation where someone's identity is needed to be proven can be integrated into the platform, as long as a developer makes an application for it. FeliCa, then, is not only out to replace credit cards, but everything else in your wallet too.

Carlo Longino is a writer and analyst that follows the mobile industry. He’s co-editor of MobHappy, and also an analyst for Techdirt. He can be reached at carlo@mobhappy.com.

Read more Airtime. The column appears every Tuesday on Gizmodo.

The 3D browser by NTT [AkihabaraNews via i4u]