Alongside the steady progress of Moore’s Law, we sometimes see step function changes in price-performance ratios of IT. One such change has produced ultra-low-cost computing (ULCC) over the past several years. In essence, this wraps computing peripherals around a cellphone hardware core; producing a computer for just a few tens of dollars. ULCC enables computing to reach applications and users that other computers cannot, and it is spinning out in multiple directions. In this column, we will focus on one ULCC development—Raspberry Pi—and one particular direction: use in developing countries.
Raspberry Pi was created to address the decline in numbers and skill levels of school leavers applying to study computer science. Eben Upton, founder of the non-profit Raspberry Pi Foundation, was an admissions tutor for Cambridge University. He felt university applicants lacked the experience of experimenting and programming that was common in the 1980s and 1990s.
In those earlier decades, the flashing cursor interface made users creative from the start and provided a direct opportunity for programming. But steady commoditization had led IT to become closed and locked down. For modern schoolkids, hardware meant sealed units of brushed aluminum and glossy plastic; software meant apps that were to be used but not understood.
Pi was therefore a “back to the future” idea; trying to open up IT in all possible ways, and to get young people tinkering with all aspects of the technology. To do that, it needed two things: a very low price (targeting $35 for the complete Model B and $25 for the cut-down Model A) and an open design.
In hardware terms, low cost was possible thanks to advances in integration that have effectively shrunk all the components of a desktop computer into a single silicon chip. The Raspberry Pi is based around a 700MHz ARM11 system on chip (SOC) with a powerful graphics co-processor. Typically this sort of processor was used in cellphones five years ago. Stacked on top of the processor is 512MB RAM on Model B and 256MB on Model A.
The SOC drives a HDMI output allowing high-definition display on the latest screens, and composite video out as used for monitors since the 1980s. The Model B has two USB sockets and an Ethernet connection, requiring less than 5W of power, with Model A (having only one USB socket) requiring 2W. Power is provided via a micro USB connector, allowing cellphone chargers or any 5V supply (including batteries, solar panels, or other renewable sources) to be used.
Apart from the graphics processor, which is propriety to Broadcom, the Raspberry Pi is completely open source, which helps to keep costs down. From the circuit schematics to the applications and the operating system, anyone can examine and contribute online. The Foundation provides a version of Debian Linux that presents users with a basic text login rather than a slick GUI by default, with the entire operating system and user files stored on a swappable SD card.
Users can “see into” the underlying mechanics of the technology—quite literally since the basic Pi comes as just a case-less board. Users must build their own computer by connecting monitor, keyboard, mouse, and other peripherals to the motherboard; and load their own operating system and applications if they want to word process, web browse, and so forth. Open design and an open innovation approach have encouraged a whole set of further developments, such as a hardware interface (Pi-Face) that can connect the computer to a variety of sensors and peripherals.
But can Pi break out of its computer science teaching and design project ghetto to address the needs of developing countries?
ULCC’s promise is that it will match the explosion in digital communications with an explosion in data processing capacity—bringing the “I” of information and communication technologies to the mass of the world’s population just as cellphones are bringing the “C.”
As we noted in a recent blog,1 three application opportunities come to mind:
- Micro-enterprise and household computing: providing access to standard computing applications for the individual enterprise and household. Add a mobile Internet connection, and we might finally move beyond the “telecenter” model of community computing, to something much more integral to the lives of those on lowest incomes.
- Technical education: as noted earlier, the prime motivation behind Pi was to reignite interest in computing as a subject among schoolchildren. There is a great thirst for IT education in schools, colleges, and universities in developing countries but budgetary constraints are a major barrier. Pi might help to overcome those but also allow students to open the box and play about much more, learning how IT works. A few projects have begun, such as the recent equipping of a village school in Cameroon.3
- Data collection and automation applications: there’s a trickle of new electronic applications for development—smart motor controllers that save power and extend motor life, low-cost health monitors, water quality and climate change measurement devices, field-based agricultural sensors. Raspberry Pi might turn that trickle into a steady stream.
At root, though, the transformational promise of ULCC may not be about a specific application, but about a new approach to innovation and design. The spread of computing applications for the poor has been limited by design-reality gaps. Applications are developed by those external to poor communities, with designs that mismatch local realities, and thus often fail.
The spread of computing applications for the poor has been limited by design-reality gaps.
If ULCCs can become widespread, they can enable a new computing design paradigm: grassroots innovation in which local users are also designers; creating designs that match local needs, resources, and context. ULCC will also allow a new model of collaborative IT innovation: working alongside base-of-the-pyramid users. Large firms, university departments, and social enterprises could now afford rapid, mass prototyping—trying out and iterating quickly through many different models until they find one that works.
But will this development opportunity be realized? Recent history is littered with the skeletons of low-cost computing-for-development failures such as the People’s PC and the Simputer.a Question marks continue to be raised over the One-Laptop-Per-Child initiative.4 So will ULCC and Raspberry Pi be any different?
Drawing on recent work analyzing what makes IT innovations scale in low-income markets,2 we can identify three domains that need to function effectively:
Supply factors. Per-computer costs in the Cameroon school were above $300, nearly three-quarters of which was the cost of monitors and monitor HDMI/VGA convertors. Sourcing or development of low-cost monitors will be a necessity if ULCC is to be the basis for PCs as opposed to the controllers/sensors of the third opportunity—data collection and automatic applications—described earlier.
Production and local distribution and marketing will also need to be addressed; something on which the Raspberry Pi Foundation is playing catchup. It was formed as a spare-time, pro bono initiative that expected to ship 10,000 units in total, and which has struggled to meet demand in the global North that is already around the one-million mark; leave alone thoughts about addressing the global South. Perhaps the best hope is that it might, as the One-Laptop-Per-Child initiative did, induce copycat private manufacturers to follow suit.
Demand factors. If, ironically, cost is not yet the unique selling point for ULCC, then what is? Likely, it is “tinker-ability” but that will limit demand to college and university sites in the developing world. Households, communities, even most schools do not want to tinker; they want something that will deliver useful applications. For them Pi in its current form may well be underdesigned and underbundled: their demand is for something more like a current smartphone or netbook. Ultra-low-cost computing will only find a major market if it can deliver those types of devices at significantly lower cost.
The alternative is, once again, the data collection and automation applications opportunity: the “gizmo route” that would find a killer application for the developing world in an electronic device built around ULCC; a device that would meet an important need of low-income communities. If that does emerge, it will come most likely from the grassroots or collaborative innovation approaches described.
Contextual factors. Even if these supply and demand factors can be met, ULCC in developing countries faces other challenges. Poor access to electricity probably is not one. This remains a major computing-for-development stumbling block: it tripped up efforts to use biometrics in the 2013 Kenyan elections. But ULCC devices’ very low power consumption means they can be used much more easily in off-grid environments.
The limited skill infrastructure is much more problematic. Underbundling demands greater skills at the point of implementation, and ongoing literacy and IT skill deficits remain significant generic barriers to computing. Yet again, packaging ultra-low-cost computing into realtively simple electronic devices may be the answer.
Conclusion
In sum, ultra-low-cost computing offers a good amount of promise for development but that is true of most IT innovations. The core issue is what lies between that promise and widespread use. At the moment, the challenges to scaling ULCC for people at the base-of-the-pyramid look daunting. Raspberry Pi, at least, may remain a tertiary education niche product with a very few secondary education implementations.
One cannot help remembering, though, that mobile telephony looked very much like this at the end of the 1990s. Firms saw no demand among low-income users and had no plans to develop that market. Only after a few innovators stepped in—including nonprofits and international donors—was there a demonstration effect that induced larger players to enter.
We will wait and see if there is any such ULCC demonstration effect, either for computers or for electronic devices. At present, this is a blank canvas waiting to be painted, held back by our need to reconceptualize; to rethink what is possible and what is feasible and what is desirable in a world of very cheap computing power.
In part the next chapter will depend on many of those within the ACM community: Can we reconceptualize and develop new ideas, initiatives, and partnerships that will fulfill ULCC’s socioeconomic development potential? Over to you…
Join the Discussion (0)
Become a Member or Sign In to Post a Comment