The theoretical steam-powered analytical engine of the 1830s is set to become a modern reality, writes KARLIN LILLINGTON

A UNIQUE project to build a lorry-sized computer designed nearly two centuries ago is gaining steam. Literally, that is. If all goes well, John Graham-Cumming will build a steam-powered analytical engine thought up by pioneer Charles Babbage in the 1830s, a man considered to be the “father” of computing.

Babbage, an eccentric and difficult but brilliant personality with a love of mathematics and engineering, invented plans for a number of devices recognised as forerunners of the modern computer but many were never built in his lifetime.

"He was 100 years ahead of his time," says computer scientist Graham-Cumming, the author of the Geek Atlas, a guide to places around the world that hold a special attraction for geeks. Graham-Cumming also successfully campaigned last year to have the British government apologise for its shabby treatment of computing genius Alan Turing, whose cryptographic capabilities were pivotal to the British war effort but who was stripped of his security clearances because he was gay. Turing eventually took his own life.

Graham-Cumming says the inspiration to build the analytical engine comes from having seen the working model of Babbage’s difference engine number two, constructed from Babbage’s original plans by the Science Museum in London in the 1990s.

“I saw it demonstrated, and to realise these machines were achievable was a pretty big deal. Babbage designed it within the tolerances of the time. And I think, like every computer scientist, I have a love of seeing something like this actually built and made to work. The analytical engine is just so clearly a computer in architecture.”

The difference engine, which also has a primitive working printer similar to a printing press, is able to do complex calculations that would have taken human mathematicians thousands of hours. But it is a far more basic machine than the analytical engine, which qualifies as a proper computer because of three things, says Graham-Cumming.

“First, it’s completely automatic – it runs on steam rather than having to be hand-operated like the difference engine. Second, it is programmable using punch cards. And third, it can look at something in its own memory and make decisions based on it. That makes it a computer.”

The analytical engine’s punch cards were based on the cards used in the just previously invented Jacquard loom to automate the weaving of cloth in the Industrial Revolution. Babbage was well aware of this invention and actually referred to his proposed analytical engine punch cards as Jacquard cards, says Graham-Cumming.

The engine’s memory is based on rods with wheels, numbered from zero to nine and stacked to get combinations of digits.

“There is a great comment from Babbage at the time, where he said that, once the analytical engine was built, it would guide the direction of science forever. He liked a bit of hyperbole,” says Graham-Cumming with a laugh.

But Babbage was probably right. The British government was so intrigued with the difference engine that it threw massive amounts of money at the project – the equivalent of the cost of two battleships. The difference engine was to calculate maths tables that could be used for navigation and astronomy, both important to a seafaring nation during the reign of the British Empire.

However, the difference engine was never built due to a combination of “personality and money”, says Graham-Cumming. Babbage could be outrageously difficult and, part way through the project, he announced that it was not worth building that particular engine but instead the government should let him build another, an approach which exasperated the government and many supporters of the project.

“I have great sympathy for the guy because he just seems to be a typical computer person who had no ability to talk to politicians,” he says. “He was a bit of a difficult character.”

At the same time, Babbage held salons that attracted many of the prominent society and science people of the day. The close collaborator on his inventions and another central figure at his salons was mathematical genius Ada Lovelace, the daughter of poet Lord Byron and for whom the computing language Ada is named.

Graham-Cumming says building the analytical engine will likely be more difficult than building the difference engine. Although nobody knew whether the difference engine would work until it was actually constructed in the 1990s, a complete set of plans was available to the Science Museum engineering team.

With the analytical engine, there are numerous sets of plans. “The problem is that Babbage was a bit like any computer person – he kept fiddling with it, adding memory, etc,” says Graham-Cumming.

Babbage gave numbers to each version of plans. The most complete plans are probably Plan 28 and Plan 28A, Graham-Cumming says, and he hopes that research will show that a combined version of the plans will produce a working analytical engine.

As technology has advanced since the 1990s, at least this time the plans also can be tested as a computer model before the actual engine is built.

To get the project done, Graham-Cumming has launched a project called Plan 28 (www.plan28.org) to raise the money and the help that will be needed to build the engine. He initially thought he would campaign to get 50,000 people to give $10, £10 or €10 on the basis that it would probably cost close to £500,000 (€588,000) to get the project started. That fund would go towards construction after an initial period of research, mostly at the Science Museum in London which holds most of Babbage’s original papers and which also is aware of the skills needed to build a working model of a Babbage creation.

But since Graham-Cumming announced Plan 28 a few months ago, he says he has been overwhelmed with much larger offers than expected. Two universities said they already have some funding for research that would dovetail with the project. He is hoping to reach the initial pledge amount by January 31st next at which point he will set up an official non-profit organisation to take the project forward.

Graham-Cumming expects the final total to build the engine might exceed £1,000,000, and is hoping corporate support, especially from the information technology industry which can fully appreciate Babbage’s genius, will bring the project to fruition. “The analytical engine is a very physical object that embodies the thinking of Babbage’s time. It’s inspirational and a sort of monument to computing.”