Article - Issue 44, September 2010
Chris Mairs FREng
A Sound Communicator
Chris Mairs FREng
As a software engineer and a founding member of a successful telecommunications software company, Chris Mairs understands the importance of communications, which is why, being blind himself, he is frustrated by the obstacles that technology imposes on vision impaired users of modern systems. He talked to Michael Kenward about his career in business and his international sporting achievements.
Had it not been for the early signs of his declining eyesight, Chris Mairs might have been lost to engineering. His inclination had been to follow his father into accountancy but he abandoned that idea when he realised that his eyesight wouldn’t be able to deal with reviewing rows and columns of figures. However, he did not entirely give up on numbers and went to Cambridge University to study for a degree in mathematics before switching to computer science.
The voice technology company he later helped to set up, Metaswitch Networks, has revenues of more than $130 million a year and employs nearly 600 people, around a half of them engineers.
As Chief Technical Officer (CTO) at Metaswitch, Mairs’s expertise is essentially in software. The CTO’s role, he explains is very much about high-level strategic vision. He predicts that, over the next few years, voice will increasingly become: “just another application that runs over the data pipe. So what I need to think about is what business opportunities can we facilitate by evolving our technology? Where should we be investing our development resources?”
Driven by curiosity
Like many software engineers it was hardware that first sparked Mairs’s interest in engineering. As a youngster, he took bicycles and typewriters apart, “always ending up with the extra couple of widgets that I couldn’t work out where they were supposed to go back in but I really enjoyed it.”
Mairs first worked with a computer before going to Cambridge. A general studies course at school involved writing software in an early programming language called Fortran to run on an ICL mainframe computer at the local polytechnic in Nottingham. His first encounter with computing as a way of earning money came when a friend talked enthusiastically about the wages that went with jobs at IBM. Mairs contacted the computer company and got a holiday job. “The money wasn’t as fantastic as he’d said, but it was very interesting,” says Mairs. He worked on converting British Leyland, the car maker, to IBM mainframes. “It was a real work-hard, play-hard project. We worked until two o’clock in the morning on a regular basis but I really got the bug.”
Mairs continued to work for IBM during university vacations. In his third year at Cambridge, he switched from maths to computer science. “It was a one-year course in those days”. This was in the late 1970s. “They couldn’t really think of enough to fill three years,” Mairs jokes.
On graduation, Mairs resisted exhortations to do a PhD. “They were quite keen for me to stay on but I had never really wanted to be an academic. I had always preferred to get on and do real stuff.” He took a full-time job as a systems programmer at IBM’s UK headquarters where there was the largest machine room in Europe. “It was the size of a football pitch,” says Mairs, and dealt with such mission critical activities as IBM’s own order fulfilment, and, for its time, an extraordinarily sophisticated database for tracking all bugs in all IBM products worldwide.
In 1982, seven IBM engineers decided to start a new company called Data Connection. They recruited Mairs as the first employee and lead architect for the company’s ‘connectivity’ software that would allow minicomputers from other manufacturers – as Mairs points out, this was before PCs existed – to communicate with IBM’s machines. This meant that point-of-sale systems, for example, could communicate with IBM mainframes. While the original idea had been to allow other systems to talk to IBM’s machines, it was the computer giant itself that bought more of the product than anyone else. IBM is still a big customer for the same software.
The transition from engineering body shop, undertaking ‘work for hire’, to a product company was crucial to Data Connection’s commercial success. “Early on I felt that we ought to be able to create communications code that was portable across different hardware platforms, and different operating systems,” he explains. “Historically that is not what people had done. Every time someone produced a new piece of hardware, they wrote the software from scratch.” Thanks to Mairs’s blueprint, the company has been able to put its software products onto as many as 300 different combinations of hardware and operating system. Over the years, the product line has evolved to include many of the core protocols that control the flow of data through the internet, with arcane names such as Multi Protocol Label Switching (MPLS) and Session Initiation Protocol (SIP).
Mairs explains: “The portable framework allowed us to sell essentially the same software over and over again to all the big players in the industry. If you look at the top 10 telecommunications equipment manufacturers, they have all got our protocol software somewhere in their product lines.”
A new division which Mairs led worked on application sharing and early collaborative software for use alongside audio/video conferencing. Mairs recalls an early attempt to sell the idea to Microsoft. Very interesting, was the response, we are working on that too. A year on, Data Connection had moved beyond its own prototypes and had something to show for its efforts while at Microsoft they were still writing the specifications. “By then they realised that it was urgent so they bought it from us.”
Sold to the staff
It takes bright engineers to write software that meets the demands of telecoms companies and network operators. They insist on reliability and systems that suppliers will support for some years. This is difficult if there is a constant turnover of staff, which is all too easy in an industry as fast moving as information and communications technologies. Data Connection took an imaginative step as a way of retaining its staff: it gave them the company. An Employee Benefit Trust was founded to hold the company on behalf of all employees.
During the past 10 years, as CTO, Mairs has played a key role in the company’s transition from supplying software to equipment vendors to becoming a substantial equipment vendor in its own right, rebranded as Metaswitch Networks. This has been a bold step, embracing new skills for the company such as hardware design, manufacturing and distribution. However, as Mairs says: “In this business today, the real crown jewels are the software, and we already had shed loads of that. The trick was to identify the very best off-the-shelf hardware and pull together a team of very smart engineers to enhance that and integrate it with much of our existing portable software.”
The voice switching solutions the company now offers, bridging legacy networks into the new world of packet based communications, are recognised by service providers, like AT&T and BT, as world leaders in their class.
Until recently, the company had no external shareholders. Unlike many tech start-ups, Metaswitch had avoided taking money from venture capitalists and was doing well. In order to move to the next level, external funding was needed and Francisco Partners and Sequoia Capital became investors in 2008.
The evolution of Metaswitch has changed the roles of Mairs and his colleagues. For example, in the early days, the company’s engineers, including Mairs, did the selling. The way to persuade potential customers to buy was to talk engineer to engineer. “That was how I got to learn about business,” says Mairs. Now, Metaswitch has a dedicated sales team that can talk the right language to service providers, where there are many issues beyond the simple excellence of the engineering.
The tie up with investors has already raised Metaswitch’s visibility in Silicon Valley, the nerve centre of the world’s ITC business. In July, Mairs decamped from the company’s headquarters in Enfield and moved a small group of engineers to Silicon Valley, a few miles south of San Francisco.
Chris Mairs takes part in the European Disabled Championships in Milan, 2000.
While Mairs has been with Metaswitch since it opened for business, there is a gap in his employment record. Throughout his career, he has had to live with steadily failing eyesight. After 15 years, he decided that his vision had declined so much that a ‘gap year’ was in order, or to be more precise, five years. “It was nothing to do with unhappiness with the company or anything like that. I didn’t want to get to the age of 55 and have gone completely blind and missed out on the opportunity to do some other stuff. I wanted to do some travelling. I wanted to get more involved in waterskiing which was my passion at the time.”
Waterskiing was just one of the sporting activities that Mairs pursued at an international standard. Downhill skiing was another, until he hung up his skis after the World Disabled Championships in Winter Park Colorado in 1990. “At the end of the downhill there, one of my other friends, who was also partially sighted and I remarked with much relief and some surprise to one another : ‘We might not have won, but we are still alive! It’s time to retire.’”
Waterskiing was another matter. “Waterskiing is fantastically exhilarating. It is something that, as a disabled person, you can get a lot out of” he enthuses. When Mairs took it up, though, it wasn’t as exciting as the able bodied version. “I was frustrated at the time that the blind slalom wasn’t a very good simulation of the able-bodied slalom.” So he brought his engineering skills to the challenge.
Mairs set out to create a technical solution. By taking the sensor out of a laser printer and the sound generator from a car anti-theft alarm, along with a cheap microprocessor, he came up with an inexpensive ‘audio slalom’. Instead of weaving around a series of buoys, blind waterskiers are judged by a device on the towing boat that measures the angle of the rope. As Mairs explains it: “Based on the length and angle of the rope and knowledge of the boat’s speed, it is fairly basic geometry to work out where the skier actually is in terms of how far down the course have they gone, how wide have they gone. You can essentially construct an audio equivalent of the buoys and then give the skier a very loud signal that they are now going around the buoy.”
Developing technology for disabled people is one of Mairs’s enthusiasms outside work. He has set up a charity to work on various ideas to get around what he sees as the obstacles that technology puts in their way. “One of my bugbears is the way in which the evolution of the digital economy actually excludes disabled people.” You would think that all this technology would make life easier for disabled people, he adds: “but far from it. The reality is that technology moves ahead much faster than accessibility to the technology.”
“Technology takes a big leap forward and that disenfranchises people with sensory impairments, particularly excluding access to the digital economy. For example, blind access to the internet is a nightmare, he says. The move to text messaging on mobile phones also excluded blind people until speech synthesis could run on the mobile handset due to improvements in processor performance.
There have been attempts to create accessibility standards, says Mairs, but then someone comes up with new ways of doing things and those standards get cast aside in the rush to add, say, multimedia, which means that: “as a blind person, you are squeezed out again.” As Mairs points out, an ageing population means that there are more people with a disability of one form or another, including poor eyesight. Even the ‘able bodied’ can benefit from better accessibility. Mairs quotes a survey on websites and their accessibility by the Disability Rights Commission which showed that sites that were most accessible to disabled people were also easier for sighted people to use. “If it is awkward to find something then this is going to be awkward for everybody, not just for blind people.”
Nor is Mairs impressed by some of the attempts to make products more ‘disabled friendly’. For example, while Apple has been praised for adding speech synthesis to its operating system, the user interface, says Mairs, “is still essentially a visual paradigm with some attempt at speaking the visual interface”. He is hopeful for future technological solutions but realises that such developments may not be driven by market forces alone. “You have got to provide a combination of sticks and carrots - the right legislative framework and appropriate financial incentives,” he believes.
For all his concerns about technological exclusion Mairs is still convinced that there are definitely roles for blind people within engineering organisations. “Despite the negative things that I have said about technology, it is good enough to allow blind people to become very good software developers. Accommodating blindness trains the mind in attention to detail and retention of facts and sequences – all useful skills in software development.”
While his current concern is to develop Metaswitch’s presence in Silicon Valley, when he eventually finds time to step back from his work at the company Mairs plans to take a more public stand on the issues facing the disabled community. More immediately, he also hopes to find something to replace his previous sporting pursuits. Two hip replacements may have put paid to Mairs’s waterskiing, but he plans to renew his ‘body straining’ activities. He hopes to complete a tandem cycle ride from Land’s End to John O’Groats. Somehow, it seems inevitable that Mairs will find a way to add some technology to the venture.