Meanwhile, in the showrooms, the big motorbikes that Triumph created for the company's relaunch in 1991 are in demand. 'Most bikes are discounted below the catalogue price,' said one salesman. 'But not Triumphs. The company doesn't do it - they sell so fast anyway.'
Even so, the list prices are competitive, from about pounds 6,000 to more than pounds 9,000. The mystery, to those unaware of the revolution in modern production technology, is how machines turned out in relatively tiny runs of only a few thousand can sell as cheaply as those made by giants such as Honda, Yamaha or Suzuki and even undercut BMW.
More surprising still, the total production of 8,000 in 1993 was divided between no fewer than 10 different models. At Bikerama in Hornsey, north London, they find that this variety is essential to the success story. The choice includes a sports machine, a long-distance touring model swathed in a streamlined fairing and an off-road bike with an upright riding position.
Common sense suggests that to be competitively priced you have to go in for mass production as understood by Henry Ford, leaving niche markets to those able to pay the difference. The reason Triumph has been able to square the circle is that it was equipped from scratch in the era of computerised manufacturing.
The company is an industrial phoenix that shares little but the name with the undertaking run by the Meriden co-operative of the Seventies - itself the last-gasp incarnation of a manufacturer which had made 250,000 machines a year as recently as the Sixties. John Bloor, a publicity-shy Midlands property developer, bought the name in 1983 and put in pounds 70m of his own money, restarting on a new site with a new range of designs.
In any previous period, someone contemplating reviving the British motorcycle industry would start with a strategic choice: either find the funds to create a tailor-made automatic production line for mass sales, or offer hand-crafted motorbikes, putting resources instead into paying relatively high numbers of skilled operators.
Traditionally, short-batch production has been slow and expensive. A part would make a laborious journey around the factory as it underwent the necessary cutting, grinding and drilling operations. The received wisdom was that machines should never stand idle. Consequently, parts could sit around for a week awaiting their turn on a lathe or a mill. Progress-chasers and expeditors ran from end to end of the production process trying to speed things up. The main bottleneck was caused by resetting time, adapting non-specialised machines for different orders; this could take eight hours or more.
Only a minority of Mr Bloor's team came from within the motorcycle industry. Alan Hurd, the manufacturing engineer, had previously been with Unipart. These outsiders brought with them a knowledge of the new systems which in the Eighties were transforming all manufacturing processes.
The Japanese had largely abandoned Henry Ford's production line in favour of an approach devised in the Soviet Union in the Thirties. This is called 'group technology' or 'cell manufacture': groups of components are made from start to finish in a 'cell' of machine tools, rather than passing from department to department.
Mr Hurd adopted this approach and also chose a Japanese manufacturer of machine tools: Enshu, which also supplies Yamaha motorcycles. These were off-the-peg machining centres of a type that can as easily be programmed to turn out food mixers as motorbike engines. They are descendants of the numerically controlled machine tools invented in 1952 at Massachusetts Institute of Technology, preset originally with computer-generated punched tape to follow a pattern of cutting and drilling without using expensive jigs. Numerical control was first devised to ensure consistency, but its most revolutionary implication was to speed up resetting.
'We can change one of these machines over in as little as three minutes, where it would take hours for an operator on an old-fashioned machine tool,' Mr Hurd says. His teams can switch the whole production line to manufacture a different- sized engine in half an hour.
The big, liquid-cooled, three-
cylinder engines are Triumph's unique selling point, distinctive because of their unmistakable growl. Seven of the 10 different models share the same 900cc version. Minor production changes reduce the engine size for a 750cc engine, and there are two four-cylinder models.
'There are really three elements in the bikes,' Mr Hurd says. 'There's the engine, the heart if you like, the chassis, which is the bones, and the bodywork, the skin. Our range is based on a modular approach, which means a number of common components can be put together in different combinations. The motorcycle business is a fashion business. Our approach meant that when we saw one segment of the market growing we were able to bring in a completely different style of bike, the Tiger, at short notice, using the same engine in a slightly modified form.'
Much of British engineering, notoriously, was liquidated in the meltdown of the early Eighties - including most of what was left of the motorcycle industry. But even after the factories were bulldozed the dealers and enthusiasts remained. Triumph takes pains to read the signals from this sophisticated and tradition-conscious market; they run factory visits and have made small changes in the bikes in response to the feedback.
But it would be rash to believe that a combination of engineering tradition and new technology can usher in a new age of British manufacturing. Triumph is different from Rover chiefly in having an investor prepared to pass up short- term returns. 'We have to take a long-term view in this business,' Mr Hurd says. 'We couldn't have done it through the banks - they wouldn't have stuck with it.'