Endaga

» see endaga.com as it lived, on the wayback machine

~ the obit ~

On 5 October 2015 Endaga's founders said they were joining Facebook and that the business was being wound down. They would not say what they would be working on there, and they would not say what would happen to the networks already running. For a company whose customers were villages with no other coverage, that second silence was the loud one.

Endaga came out of UC Berkeley's Technology and Infrastructure for Emerging Regions group, which had spent years on the question of why cellular coverage stops where it does. The answer is not technical. A tower costs what it costs, and a carrier builds one where the subscribers will pay that back. Below some population density the arithmetic simply fails, and about a billion people live below it.

The product was a gray box about the size of a microwave oven, with solar panels, containing the radio hardware and the software to run a small GSM network on its own. Mount it on a tree. It handled calls and texts inside the village, connected outward over whatever backhaul existed, and came with billing built in so that a local operator could charge for service and keep the money. The company incorporated in 2014 with $1.2 million of seed funding behind it.

~ the rap sheet ~

~ the economics ~

The interesting claim was never the radio. Community cellular hardware existed before Endaga and the open-source GSM stacks it leaned on were years old. What Endaga packaged was the business: a village-scale operator who buys a box, sells airtime to neighbors at a price those neighbors can pay, and services the loan out of the revenue. The research deployment that started it all went up in a village in Papua, Indonesia, in 2013, and ran as a going concern with its local operator keeping the proceeds.

By Berkeley's own summary, thousands of people got cellular access this way. That is a small number against the billion in the pitch deck and a large one for a company that was two years old and had raised less than most seed rounds.

~ what joining facebook meant ~

Facebook in 2015 was pushing hard on connectivity in exactly these markets, with Internet.org and its satellite partnerships, so hiring the three people who had made rural cellular work commercially was not a strange move. The founders joined the general engineering team rather than Internet.org itself. Facebook announced OpenCellular, its own open-source cellular access platform, the following year.

Endaga's own site kept the acquisition banner across the top of its homepage for years after there was no company: "The Endaga team is joining Facebook. Read more." The archive still has it in March 2016, still in November 2017. By November 2018 the domain was a registrar's placeholder saying endaga.com is a totally awesome idea still being worked on.

~ the distance, which is not the usual kind ~

My own first network was a 14.4k modem in Turkey in 1998, which felt slow and was not. It was a service sold to me by a company that had already decided my street was worth wiring. Nothing about my connectivity has ever depended on whether anyone in my neighborhood could personally operate the equipment, and that gap is the whole subject here.

Kurtis Heimerl and Shaddi Hasan went into academia afterward and kept working on community cellular networks, which is where this research came from in the first place. The first network, the one in Papua, kept running after the company that sold it had been wound up.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »