Outback technology

The western highlands of New Guinea, in the Indonesian province of Papua, are home to one of the planet’s most remote inhabited regions. In the fall of 2012, EECS doctoral students Kurtis Heimerl and Shaddi Hasan headed there to find a village called Desa.

“It takes two or three jet hops to get to Jayapura, Papua’s capital,” says Heimerl. “You wait overnight for a plane to Wamena” — the largest town in the highlands — “and if you’re lucky, you can find a taxi when you get there.” The taxi is a truck, fare U.S. $20 to stand in the crowded back, $25 to sit in the cab, but either way you get out to push through the muddy stretches on the one-lane track. After four hours or so, barring breakdown, you’re in Desa.

Heimerl and Hasan are members of the Technology and Infrastructure for Emerging Regions research group (TIER), directed by EECS professor Eric Brewer. Their purpose in Desa was to construct a cellular phone network that would be low-cost, sustainable and, within a short time, locally owned and operated, with all profits staying in the community. Desa’s cellular system was the test case for TIER’s Village Base Station project.

Heimerl, who was recently selected as one of “35 Innovators Under 35” by the MIT Technology Review, grew up in a small town in Alaska, a state where accessibility can be an issue by dogsled, much less cellphone. At Berkeley in 2009, while pondering a subject for his Ph.D. thesis, he came upon OpenBTS software, which combined essential components of a 2G cellular base station in a laptop computer. Add a power supply and a tower — which might be no more high-tech than a tall tree — and you’ve got a base station, for a cost of only a few thousand dollars.

Men in tree to install the Village Base Station (photo courtesy of Kurtis Heimerl)

Heimerl’s inspiration, the Village Base Station project, was a perfect fit for TIER’s goal of “bringing the information-technology revolution to the masses in the world’s developing regions.” It also meshed nicely with Brewer’s educational philosophy.

A legendary innovator, Brewer has never hesitated to tackle daunting ideas. In 1996, while an assistant professor at Berkeley, he and one of his students co-founded the Inktomi Corporation; the company, whose structure prefigured cloud computing, soon rose into the NASDAQ 100. When it comes to teaching, Brewer says, “My job is to give students the permission to do what they want to do. We make them find out if it’s possible.”

In 2010, Brewer and Heimerl presented the Village Base Station concept at an Association for Computing Machinery workshop on networked systems for developing regions, in San Francisco. A subsequent talk by Heimerl about OpenBTS 2G systems made it onto the web and attracted the attention of the owner of WamenaCom, an Internet service provider in Papua whose services are used by Western missionary agencies. In Papua, Heimerl says, “the missionary agencies are much more community-minded than folks might immediately assume.”

WamenaCom invited TIER to work with one of their clients, a Methodist missionary school in Desa, run for the last 10 years by an American couple. WamenaCom and the school would eventually divide the modest income from the cell network, with WamenaCom providing technical support and the school chipping in electricity, administrative help and security.

Launching TIER
With Inktomi’s success, Brewer had been thinking of setting up a foundation, and asked himself, “What does it take to improve health and prosperity in a developing nation — or even a single village?”

Brewer says, “Most people will tell you the answer is governance and macroeconomics — and they’re right. But just as important, and often overlooked, is technology.”

After the dot-com bubble popped in 2003, Brewer and his partners sold Inktomi to Yahoo for rather less than its former stellar worth. His determination to bring technology to emerging regions persisted, but instead of a foundation to bestow grants on worthy projects, he created a research center that would come up with worthy projects foundations couldn’t resist.

One of TIER’s first projects was to devise a long-distance WiFi system — an oxymoron at the time, with WiFi protocol limiting range to a few hundred meters — by writing novel software to optimize low-cost, rugged, off-the-shelf hardware. Learning that the Aravind Eye Hospital in Theni, India, hoped to use telemedicine to reach patients in rural areas, where poverty and distance often kept those with eye problems from seeing a doctor until it was too late to prevent blindness, TIER sent graduate student Sonesh Surana to meet with the hospital staff.

Just 17 months after that one-day meeting, a team of four TIER graduate students, working with local vendors, had established WiFi teleconference links with five vision centers, each run by a trained nurse and each serving over 4,500 patients a month. By then, network operations were already in the hands of the Aravind staff.

In 2004, Theni was one of five hospitals in Tamil Nadu’s renowned Aravind Eye Care System. On a return visit this past spring, Brewer found that Aravind had expanded to 11 hospitals with telemedicine connections to 40 remote vision clinics, treating or referring a total of 310,000 patients a year for endemic problems such as cataracts, diabetic retinopathy and glaucoma.

What he hadn’t expected, he says, is that the ophthalmologists he spoke with felt that the rural patients may be getting “as good or better care than those who show up at the hospital. The nursing assistant has seen more of the patient than the doctor has. Two sets of eyes are better than one.”
Brewer likes to say that academics observe, engineers intervene. “It took me a while to appreciate this dichotomy,” he says, “but the result is that in TIER, we try to discover what the actual problem is before we try to fix it. That’s something engineers historically are not good at.”

TIER focuses on immediate problems, but what underlies the program is Brewer’s vision of the future. His inspiration for Inktomi in 1996 enabled banks of remote servers to provide scalable, flexible and resilient storage and processing to any computer with access — what’s now called the cloud.

Just a decade later, people were toting hand-held supercomputers, a.k.a. smartphones, whose sales were about to outpace PCs. Brewer saw that what it really takes to turn a smartphone into a supercomputer is access to the cloud.

“Coverage is what determines the cellphone’s value, and until now, cellular has been mostly urban,” he says. “The real digital divide is not between nations or even classes. Even in the U.S., it’s between urban areas and rural areas.” It’s a divide that TIER is determined to bridge.

Photo courtesy of Kurtis Heimerl

‘Wild West’ of telecom
Back in Desa, the base-station idea seemed not just technically possible but easy. The station was up and running on the first day, sharing the school’s Internet tower — it was indeed a tree — and power from the school’s small hydroelectric generator, which shut down at night so a jungle stream could refill the small reservoir behind the low dam.

Then came the hard part. Local cellular networks around the world constitute what Heimerl calls “the Wild West of telecom.” Desa’s telephone numbers had to be acquired from Sweden; its SIM cards were custom-made in Berkeley; and local vendors in Desa were chosen to sell usage credits. Negotiations with established carriers — who didn’t serve the region but were essential to out-of-network communications — and other complications took weeks and required more trips between Berkeley and Papua.

Although approved by local government officials, Desa’s network is technically illegal: the license to the communications band is held (although not used) by those distant carriers. Nevertheless, the Desa network opened for business in early 2013. By this summer, it was serving 349 users and had handled 45,000 local calls, 100,000 local messages and more than 270,000 outgoing and incoming out-of-network messages.

At first the calls and messages concerned mostly family and community matters, but lately customers have been using the exchange of usage credits as a new kind of informal currency. As the network grows, changes are inevitable, with potential impacts on agriculture, health and the local economy. While improvements are still contemplated, Desa works.

Meanwhile, Heimerl, now a postdoctoral researcher, and his TIER colleagues are looking to other remote regions in need of cellular service, in Afghanistan, Zimbabwe — and yes, even Alaska.

Synergistic solutions to global problems
The scope of technological solutions pursued during TIER’s first decade is wide and multifaceted, with current and past support from companies and groups such as Intel, Google and the U.S. Agency for International Development, through the Blum Center for Developing Economies, the Center for Effective Global Action and the U.S. Department of Energy’s Lawrence Berkeley National Laboratory. Projects vary from figuring out ways to use text messaging for posting to Facebook, to assessing the rural potential of so-called white space (licensed but unused cellular spectrum), to developing a novel rapid-feedback data platform called Mezuri (“measure,” in Esperanto), which could allow products and services to be redesigned in real time even as they are being deployed.