Show Hide image

The NS Profile: Mary Lou Jepsen

The pioneering designer of the first cheap laptop for the developing world, she is determined to clo

These days, LCD screens are an ines­capable part of the view: on computers, in televisions, as digital adverts, on mobile phones. But as ubiquitous as they are, the technology behind them is seldom thought of as revolutionary. However, Mary Lou Jepsen’s innovations are the exception. As chief technology officer at the non-profit organisation One Laptop Per Child (OLPC), the pioneering screen designer created the first laptop for the developing world. The XO is the cheapest, least power-hungry notebook computer ever produced, a device that may eventually prove one of the most important educational tools of its time – and for which last year Time magazine named Jepsen one of the world’s 100 most influential people.

Along the way, her design sparked a mainstream computing revolution. Jepsen’s primary intention was to “innovate at the bottom of the pyramid”, creating a simple computer for educational use in impoverished areas. But it turned out that computer users in industrialised countries also wanted inexpensive, environmentally friendly laptops. The netbook was born. Just two years after the XO was first released, nearly every major PC brand is selling an inexpensive, low-performance mini-laptop, and analysts predict sales will have reached 50 million by the end of the year. “Every time I meet with the CEO of a big laptop company, they tell me they ‘studied’ my design,” Jepsen has said. Now the chief executive of her own company, Pixel Qi, she is working on computer design developments that will potentially cut battery use in half, and is on the shortlist for the 2009 Time 100.

How did screen design bring Jepsen to this point? The key is that in any laptop, the display is the most power-hungry component as well as the most expensive to produce. When Nicholas Negroponte, founder of MIT’s Media Lab, first conceived the OLPC project in 2004, he had no intention of redesigning the laptop. His plan was to mass-produce a computer from components available, then sell them in large batches to the governments of developing countries for $100 each. But when he approached the Taiwan-based computer manufacturer Quanta for advice, Ted Chang, planning head of the Quanta Research Institute, told him to take another step back before he moved forwards. “We could point him to many contract manufacturers who knew how to assemble inexpensively, but we didn’t see that as the key challenge,” Chang explained to Willy Shih, a Harvard Business School professor who has studied the OLPC project. “OLPC needed somebody who knew how to design for low cost.”

Meanwhile, Mary Lou Jepsen was designing HDTVs for Intel. Brought up on a family farm in Connecticut, Jepsen first became involved with technology as a college student, spending the summer after her junior year test-driving submarines for the US Naval Underwater Systems Centre in her home state. By the time she joined Intel, she had already spent almost a decade running her own screen design company, MicroDisplay Corp. Before that – following a degree in art and electronic engineering at Brown University – she co-created the first holographic video system as part of her studies for a Master’s at MIT. After a few brief appointments teaching computer science, she returned to Brown for a PhD in optical sciences. The decision to return to academia was a purely practical one; she has described the qualification as nothing more than “a stupid union card”.

Even so, it was a union card she made more effort than most to get. In the final year of the degree, a cancerous tumour was found on her pituitary gland. It was removed, and within six months she had completed her PhD and founded MicroDisplay Corp. Her body now produces no hormones, and so, without a strict regimen of a dozen pills a day to keep her levels stable, she could not survive. However, there is a plus side: jet lag does not affect her – which must be a bonus, because Pixel Qi is based in both Taipei and California.

During her time at Intel she had become frustrated at producing high-spec products for “1 per cent of the population”, she says, explaining that she had begun to make plans, independently of her interview for a job at MIT’s Media Lab, to start a company designing screens for the developing world. “I decided to join with Nicholas the day after my interview,” she says. Moving from California to Boston, she became OLPC’s first employee.

The project has been mired in controversy more or less ever since. Micro­soft’s Bill Gates was one of OLPC’s earliest critics; Apple’s Steve Jobs, Craig Barrett of Intel and Michael Dell also ridiculed it. “That got us press,” Jepsen points out. Even the organisation’s most basic aim has been challenged: in 2007, Nigeria’s education minister of the time, Dr Igwe Aja-Nwachukwu, criticised the organisation, asking: “What is the sense of introducing One Laptop Per Child when they don’t have seats to sit down and learn, when they don’t have uniforms to go to school in, where they don’t have facilities?”

It is true that OLPC has never engaged with grander questions of educational infrastructure. Instead, it has concentrated on providing a resource that is arguably a lot more important than chairs: information. “The world’s information is digital,” Jepsen pointed out at ETech, the annual O’Reilly emerging technology conference, in March. “The web, the news, all of that is digital. And now . . . we have ten million books scanned. That was the last bastion of what was offline; it’s now online and accessible.”

All of that information is available to the 1.6 billion of us with internet access – almost all of whom live in developed countries, a pheno­menon known as the digital divide. According to a 2003 EU report, industrialised countries accounted for 88 per cent of internet users – and just 15 per cent of the world’s population. But there is a more astonishing figure: 97 per cent of the world’s adolescents live in developing countries. So, Jepsen told her ETech audience, “If you want to change the world . . . we need to figure out how to give these kids a chance.”

Making a laptop for the third world presented Jepsen with a completely novel set of considerations. The XO had to be more than ultra-cheap and durable; it had to be suitable for children whose access to electricity is intermittent, many of whom learn outdoors in hot, sunny climates. To this challenge, Negroponte says, “Mary Lou brought technical excellence and out-of-the-box thinking, fully prepared to use both to fly in the face of industry presumptions”.

Her resourceful design included open-source software and a low-performance microprocessor to keep costs down, and a flash memory – like the kind in a USB stick or MP3 player – which is cheap, low-energy, and difficult to break because it lacks moving parts. The XO was designed with antennae, letting the laptops connect to the internet by forming a “mesh network” with one another, so that, as Jepsen puts it, “if one laptop in the village is connected to the internet, they all are”.

But the most important innovations came with the laptop’s screen. Jepsen has described her approach to the XO as “designing the computer kind of backwards”. She points out that her specialism is unusual, even in that field: “I’ve found that people who design computers don’t know a lot about displays.” She designed the XO display with two modes: an “e-paper” mode for use in sunlight, with a high-resolution black-and-white image, and a colour mode, using a backlight at night or indoors. And she devised an LCD panel that can detect when an XO user is looking at a static image (such as a page of text) and switch off the main processor, thus saving electricity. This brilliantly original idea is largely why, when most laptops’ power consumption is between roughly 10 and 45 watts, the XO uses about two.

In November 2006, the first prototypes were completed. Kofi Annan had praised the project as an “expression of global solidarity”, and 50 countries had made inquiries. Jepsen’s innovative design received almost unanimous acclaim.

OLPC, on the other hand, still had its troubles. The cost of the XO was still around $175 – unthinkably cheap by most measures, but still well above the $100 price tag OLPC had promised. Then, in March 2007, Intel launched its Classmate netbook, which it sold to countries including several OLPC had been focusing on – Nigeria, Brazil and Pakistan. In the ensuing spat, Negroponte took the XO into the US market, before Intel and OLPC briefly joined forces but then abandoned the joint project in early January 2008, each accusing the other of undermining its particular laptop. Meanwhile, the numbers of XOs being shipped were nowhere near what Negroponte had originally envisaged, and the organisation’s move into the US market was not smooth. The availability of commercial netbooks inspired by Jepsen’s original design had a significant effect on the XO’s domestic sales.

That same week of January 2008, Jepsen had announced she was leaving the project. “I miss OLPC, but my job was truly done,” she says now. “I created the laptop against the odds . . . The best thing I can do to help OLPC is to continue to work where others aren’t concentrating – to develop outdoor-readable and low-power screens.”

She is still involved with the organisation. “I just met with Nicholas last week. We are in close collaboration.” And Negroponte remains bloody-mindedly cheerful about his project’s future. “OLPC is on the cusp: 1.5 million laptops have been sold, 850,000 of which are already in the hands of kids in 31 countries.” While critics point out that Intel have shipped twice as many laptops to the developing world, Jepsen supports Negroponte. “The key thing is, a million children now have laptops across the developing world that otherwise would not,” she told ETech. “A million is a good start.”

Meanwhile, Jepsen is designing new products at Pixel Qi. “We are focused on the whole pyramid, not just the bottom of the pyramid, because this is about volume,” she says. “If you can make a lot more of something, you can make it much more inexpensive.” In many ways she feels her work at Pixel Qi is not a huge move away from OLPC. “We see strong potential to use screen technology to massively drive down the overall cost of a system so that many more people can cross the digital divide,” she says.

Jepsen admits the company has in some ways been a trying venture. “Perhaps the stupidest thing I have ever done is start a company in two cities, neither of which I lived in. From Boston, in the worst economic meltdown for 80 years, I started a company with dual headquarters in San Francisco and Taipei.” And on top of the logistical difficulties, there is the economic climate, which has made it “very hard to do fundraising”.

The first proper test of Pixel Qi’s mettle will come next month with the release of its 3Qi screen, a three-mode display that picks up where the XO’s dual-mode screen left off. Jepsen also hopes to create a screen requiring no more than half the battery power of current models – which will be welcomed by commuting netbook users and children and teachers in ­areas of the world where electricity supply is minimal or unreliable. She is also planning a low-power TV for the Indian market in 2010.

Competing with LCD is about more than picture quality, she says. “If you want to make something you can ship, you must consider the manufacturing infrastructure. LCD manufacturing infrastructure is, I believe, the most expensive ever created in the consumer electronics industry. To try to make a screen that needs a different manufacturing base is a bit like trying to make a computer chip and not using silicon.”

So, you may find your daily life peppered with LCD screens for some time yet. But Jepsen’s efforts will mean that they are more environmentally friendly, cheaper to run and better to look at; and, more importantly, they should at some point be available to absolutely everybody.

Mary Lou Jepsen: the CV

1965 Born in Connecticut
1983-87 Studies electrical engineering and art at Brown University
1987-89 Studies for Master’s in holography at MIT. During her studies she co-creates the first holographic video system
1989-92 Teaching posts at the University of California San Diego, RMIT (Melbourne), Academy of Media Art (Cologne)
1993-96 Returns to Brown University to study for a PhD in optical sciences
1996 A cancerous tumour is found on her pituitary gland. Six months later, she completes her PhD
1995 Co-founds MicroDisplay Corp; remains the company’s CTO until 2003
2004 Becomes chief executive of Intel’s display division in Santa Clara, California
2005-2007 CEO of One Laptop Per Child
2008 Founds Pixel Qi in San Bruno, California, and Taipei, Taiwan

This article first appeared in the 04 May 2009 issue of the New Statesman, Flu: Everything you need to know