Connected: the scientist whose life inspires a major new film, an obscure research unit in Scotland, flying the space shuttle and the invention of the internet

Less well known is the Glasgow connection with Turing, and how the brief but brilliant life of the research centre founded in the city in his name made Scotland a global centre of the commercial appliance of science.

Now all-but forgotten, the George Square-based Turing Institute, created in 1983, pioneered new ­frontiers of artificial intelligence through research and development tie-ups with leading businesses and government agencies around the world. Peter Mowforth, 61, then a young Cambridge PhD who became a core member of the Turing team, says: "We taught Nasa to fly the space shuttle; created an early version of the web [see panel, right]; helped the European Space Agency control its satellites; helped BA maximise flight booking yields; helped Nippon Telegraph configure its networks; helped Enterprise Oil with its discovery programme, and even helped Unilever offset fluctuations in the prices of ice-cream ingredients."

The Institute also staged the world's first Robot Olympics in 1990, an international gathering of mostly non-humanoid automata.

But as with the story of Turing, who took his own life in 1954, the story of the Institute did not end happily. The technological and commercial empowerment that should have followed was never fulfilled and it went out of business in 1994, amid questions in the House of Commons and bitter recriminations between the scientists and the Scottish Development Agency quango, which seized the reins just as research for artificial intelligence (AI) was drying up.

Glasgow's link with Turing runs through Professor Donald Michie, the brilliant and charismatic mathematician and geneticist - and the only person who could beat Turing at chess at Bletchley Park, home to the Government Code and Cypher School during the Second World War.

Michie never spoke of his work there, and could only rarely be persuaded to talk about Turing. Michie's career included founding in 1967 and being chairman of Edinburgh University's Department of Machine Intelligence and Perception - a department he created without the knowledge of the university authorities.

Following on from his Bletchley Park work on Colossus, the world's first electronic digital computer, Michie was an early pioneer of machines that can learn. Under his directorship, Edinburgh's AI department grew until around 1973, most of the AI research in the UK was taking place in Scotland.

But Scotland was to lose out after 1973's Lighthill Report. Formally called Artificial Intelligence: A General Survey, it was compiled by James Lighthill for the British Science Research Council as an evaluation of academic research in AI. The report essentially said it was overrated, stating that "in no part of the field have discoveries made so far produced the major impact that was then promised". This study formed the basis of government support for AI research, which was reduced to three centres.

Mowforth says: "Someone decided that Britain only really needed three computers as there wasn't much future in it. Not for the last occasion, at a time when Scotland ruled the roost globally, the plug was pulled."

The timing was unfortunate. Faced with the decline of heavy industry, Britain's failure to invest in cutting-edge science that could prove economically ­transformative only began to be reversed in the 1980s, largely in reaction to Japanese advances in software design.

Michie, by then retired and wary of governments and universities, was inspired to create a private sector-led centre specialising in the applications of machine learning. But public support would only be forthcoming if the body was based in Glasgow, not Edinburgh. The facility - originally called the Machine Intelligence Research Unit - was established at George House, on the north side of George Square. In homage to Michie's former chess opponent, it was named the Turing Institute, although it was a not-for-profit research company with loose links to Strathclyde University.

Mowforth says: "No-one in Britain had ever really heard of Turing, largely because of the secrecy that had surrounded Bletchley Park, and also the cloud cast by the shame of his early death [he took his own life after being "chemically castrated" following an arrest for homosexuality]. But we found that among computer scientists in the US and Japan and elsewhere, he was regarded as a superstar."

Thus began the glory days of the Institute, which at one point was twice as large as any other AI research hub in Europe, employing 70 of the finest minds from around the world.

When the UK's Nimrod radar research programme was scuppered in favour of the US Awacs, a quid pro quo was to make the Turing Institute, along with other UK science centres, into an approved centre for US research partnerships, including the Office of Naval Research.

These stellar wins for Scotland were highly classified and are being revealed for the first time by the Sunday Herald today. The jewel in the crown of the Turing Institute was its work on Nasa's space shuttle, the landing system for which, Mowforth says, was "built in Scotland from girders".

He adds: "As far as I know they used our technology on every flight from the beginning to the end of the programme and it never failed."

When the shuttle was first built, he says, it was "a bit like a flying brick, with very small wings. Test pilots found it extremely difficult to fly. They were looking for an electronic pilot scheme that could land it - they had a numerical simulator to understand the physics of it, but they didn't know how to go from that to building a control system."

A scientist called Charlie Riese of Radian, a US tech client of Nasa, heard that work was going on in Scotland on so-called "inverted pendulum" experiments, in which the computer "learns" to control an object by collecting a vast number of failures, which are fed into a learning algorithm. Eventually the computer will "learn" to control the object.

"They gave us the conventional numerical simulator that explained the physics of the thing, We ran it for weeks and weeks - computers were slow then - and generated millions of simulated shuttle crashes. With the big data generated, we ended up with a computer programme that was able to fly it."

Similar work was done for Enterprise Oil on a system which would "read" geological maps and determine the optimum spots for test drilling.

In his Budget speech in March this year, Chancellor George Osborne announced a £42 million, five-year project for a new Alan Turing Institute. While bids from universities around the UK are invited, the expectation is it will be based in Manchester, Turing's birthplace, though Edinburgh's continuing excellence in big data and informatics means Scotland may win some satellite benefits.

Mowforth now champions the appliance of science to business as an evangelist for Scotland's e-commerce skills sector. And like the man after which it was named, Glasgow's Turing Institute's will one day receive its due.

Slipping through the web

But for a grim twist of fate, Glasgow's Turing Institute could have become the birthplace of the World Wide Web. One of the Institute's global affiliates was the Society for Worldwide International Financial Telecommunication (Swift) - a network of global banks, with its headquarters in Belgium - which maintains the "plumbing" via which cash is moved securely around the world.

The Turing team was subcontracted by Swift's research and development boffins, led by an American engineer called Arnaud David Rubin, to solve a simple problem: how to distribute a user manual that needed regular updating as changes were made to the system.

The Turing Institute's great inspiration was to simplify the electronic format to its barest essentials and add one essential ingredient: hyperlinks. These could transport the reader from one part of a document to another, and between documents. It went live in 1988.

The real eureka moment was the invention of virtual documents - primitive web pages - that could be written and read by computers on screen. From there it was a short step to using the Swift network to distribute and share these documents around the world as marked-up electronic pages.

Peter Mowforth comments: "It wasn't fancy science, it was a simple thing. The only content it could work with was a financial services technical user manual, not exactly kitten videos."

So why is Tim Berners-Lee, a physicist at Cern in Switzerland who produced a similar means to share electronic documents two years later, credited with being the "inventor" of the web? The answer is that the Glasgow web breakthrough was a privately funded project subject to commercial sensitivity, but Berners-Lee gave his version away for free to fellow academics.

Mowforth says: "To be fair, while we believe we were the first, Tim's was a far richer implementation than ours."

The Glasgow web might have "grown arms and legs" the way Berners-Lee's did two years later. But shortly after the system went live, Rubin was returning to New York. He was on was PanAm Flight 103, and the date was December 21, 1988. The aircraft blew up over Lockerbie, and Rubin was among 270 people killed. The personal link between Swift and the Turing Institute was broken, and the application's potential was never realised.