The "maglev" train gets its name because it levitates a few centimetres above its monorail track, held up and propelled forward by electromagnetic effects. There have been commercial "maglev" systems. Actually, only one. Until two years ago, it was Birmingham Airport's claim to world fame: a "maglev" train carried passengers between the air and rail terminals, a 90-second journey that was impressive only in its daftness.
Nowadays, the job is done by shuttle buses. They may be unglamorous and noisier, but they are cheaper. So, will Japan prove us all wrong by turning the "maglev" into our future, the one imagined by so many visionaries of 50 years ago?
"I doubt it will ever happen. There's no rationale for it," says Laurant Delgrande, a transportation analyst. But why not? "Maglev" systems are simple in principle. They work by magnetic repulsion, between magnets mounted on the moving carriage and electromagnetic systems in the rail.
First, you have to energise the rail, so that it will produce an electromagnetic field that lifts the weight of the carriage. Then, you have to move that field forwards along the track so that the carriage is both pushed from behind and pulled from in front by a "wave" of magnetic propulsion.
The advantages are that there is no friction and no moving parts. The problems: the amounts of energy required are huge, and there are safety questions, too, such as, how do you stop such a train safely? With no friction, the idea of braking becomes tricky.
Whatever happened to "maglev", one is tempted to ask - except the real question is, whatever didn't happen? To which the answer is, cheap superconductors.
For a "maglev" train to float it has to carry big superconducting magnets, which - using present technology - means keeping them cooled at around absolute zero, -273C. At that temperature (around that of liquid helium), the current keeps circulating without loss in the electromagnets on board the train, maintaining the repulsion with the track.
Above it, the superconducting effect disappears and the train stops abruptly. Most other countries have already given up on "maglev": the United States stopped development 20 years ago, and France is sticking with its conventional high-speed train, the TGV, which achieved a speed of 515.3 kph (319 mph) in 1990.
Germany is the only other nation besides Japan still developing a "maglev" train - for a 290km (180m) connection between Hamburg and Berlin, the country's two biggest cities. The government formally set up a management company for the project earlier this month, and construction is officially planned to start next year. But environmentalists and others criticise its huge cost - DM9.8bn (pounds 3.5bn).
What does Japan like about "maglev"? For a culture always in love with technology, but also with safety, the idea of mass transit by "maglev" offers a way to get rid of wheels without taking the risks of the skies.
The intention is to build a route between Tokyo and Osaka, Japan's two largest cities, a route that 360,000 people take every day. The trouble, really, is the cost.
Recently, a test track 18.4km long was built outside Tokyo. The cost: Y150bn (pounds 680m), or (if you prefer imperial measures) pounds 23m per mile. At that rate, covering the full 770km (481m) to Osaka would cost pounds 29bn. It makes the Channel Tunnel seem like a licence to print money.
Of course, the engineers argue that initial costs are high because they need costly computers and sensors to gather the test data; costs will fall in mass production. The problem is, "mass production" is unlikely to be a reality with "maglev" until the price of superconducting materials falls substantially, or the temperatures they work at rises dramatically.
That seemed like a possibility early this decade, when superconducting temperatures leapt upwards. But they have stuck around -196 C, the temperature of liquid nitrogen; and although that liquid is nowadays as cheap as milk, the refrigeration units required still burden your "maglev" train.
Add to that the huge cost of the electromagnets required in the track itself - plus the requirement that, to travel safely at speeds of 500kph (310mph), the "maglev" guideway must be flawlessly straight and flat - and you have a science and engineering problem that can only be solved by money.
The trouble is, it has already been solved, more cheaply, by old technologies. "Rail technology has advanced to the stage at which it can do most of what `maglev' was intended to do," says Tony Eastham, an engineering professor at the Hong Kong University of Science and Technology. "It is not quite as fast, but it is more affordable, and it is a proven technology," he says.
Japan's shinkansen, or bullet trains, run on old-fashioned wheels and rails, are reliable and very fast: the newest models operate at speeds up to 300kph (186mph). The fastest bullet train makes the Osaka to Tokyo trip in 2 hours 30 minutes - a "maglev" train could cut an hour off the trip. But analysts doubt the time savings would justify the cost of an entirely new system.
"Every part of the project is new and expensive," concedes Satoru Sone, an engineering professor at the University of Tokyo who is advising the railroad on the "maglev" project. He says estimates of the costs of linking Tokyo and Osaka with a "maglev" line range from Y3 trillion to Y10 trillion (pounds 13bn to pounds 45bn). But even the best approximations are a stab in the dark. "Frankly, I don't think anybody knows how much it will all cost," he says.
Barring miracles - or better science - it looks like those visions of years ago will remain just that: visions.