PCs For The People, By The People — Will The People Buy Them?

The world on the other side of the digital divide regularly reverberates with announcements of cheap PCs. In India, where there is constant talk by government and non-government organizations, the next generation Simputer, called Mobilis, has been just released with great fanfare by Encore Software together with the Ministry of Science and Technology. At US $200+, Mobilis incorporates various home grown technologies and is targeted at low-end office automation and military applications. At the same time, other companies, like Xenitis and Celetronix are introducing similarly priced low-end PCs which use more standardized components.

But are these reverberations simply empty vibrations?

Cheap PCs typically include slow processors with small memory and offer inferior graphics. Are these devices really capable or economical enough as incarnations of generic PCs? For just a little more money, a much higher performance PC can be purchased that also has a better selection of the latest standard components.

The problem is that cheap PCs are still expensive. Today they have a target price of $200-$300. This is the cost of housing per year in some developing countries. But even this price has been difficult to achieve. For example, India’s Simputer was never able to build the volumes that could bring down its prices. Similarly, Mobilis will have to sell 50,000 to 100,000 (and by some estimates 1,000,000) units to approach its $200-300 target price range. That volume may be hard to reach since many potential purchasers believe that it is wiser simply to buy a full capability, standard PC from the start.

“Focus on child-like simplicity, radical hardware integration, plus content and software built by open source collaboration.”

The second and bigger problem is that the cheap PCs are still PCs. Most real PC purchasers in developing countries are already PC literate and prefer full capability machines, even if they are more expensive — ironically, often only marginally.

The current crop of cheap PCs are intended for markets that are filled with relatively unsophisticated consumers of information processing devices. But cheap PCs do little to solve the problems of computer usage in the native languages of these target purchasers. Unsolved localization problems inhibit widespread PC adoption just as much as price does. Problems include supporting interface tasks like keyboard input in local scripts and searching and sorting of data encoded in cross-language, multi-byte, non-Latin character sets. Today’s cheap PCs are simply dumbed down versions of the expensive, complex, and fragile PC which was designed for professional office or sophisticated home environments. The cheap versions continue to require all the complex human interfaces built for users already possessing a basic threshold of information processing literacy.

What’s needed is a radically different design point!

A people’s PC must have high utility but at the same time be simple to use in multiple local languages. Plus the target system must be rugged and be able to survive in a semi-outdoors environment — an environment that will be inevitably too hot, too wet, or too dusty. Think about how a people’s PC across the digital divide must function in an open classroom or in the market center of a farm village at 110°F with intermittent electricity and free-roaming animals. To be reliable in such an environment, it must have few or no moving parts.

The right design point is not a “people’s PC”, but a “people’s appliance”. For the vast majority of users, especially school children, an appliance with embedded functionality could be designed to be extremely easy to use. A true local language interface for non-Latin scripts can be achieved by using a touch screen with menus and icons dynamically rendered in local fonts as needed. A people’s appliance must also have compelling content, such as official school board curriculum material or voting registration lists or local land record information. It must also have simple connectivity options.

The design of a people’s appliance should not repeat the infinitely capable but dauntingly complex character of the PC. It should instead be built from a basic platform that can be adapted to the specific needs of its target users. A successful paradigm for this kind of appliance already exists - the modern game console.

For example, in a fashion similar to the distribution of content for game systems, a school appliance intended to deliver educational content could employ a platform engine that is then customized by appropriate curriculum cartridges or DVDs. Human interface tasks, like input and selection, can be serviced by a simple touch screen interface rendered in the local language. Targeting a different application, a village land records appliance would use this same platform engine but would be customized with an application-specific data cartridge or DVD.

Given sufficiently large scale System-on-Chip (SoC) integration, extremely reliable and capable platforms can be built cheaply. Many of the most important basic tasks in a village do not require high processing power or sophisticated graphics handling. A variety of advanced SoCs that run Linux exist today in production quantities and are more than capable of supporting many fundamental applications like simple curriculum delivery and basic records management.

Open source software and tools together with open content developed in local languages can provide the information layer needed to fulfill the promise of the platform engine.

In theory at least, an example of such a simple appliance that could support a surprisingly large variety of applications, is the children’s drawing computer called Pixter, available from Fisher-Price for less than US $80 retail. This product offers an interesting model for what is possible in terms of scale of integration, ease-of-use, as well as deployability. Turning such a device into a people’s appliance would require redefining some of the I/O interfaces, introducing font rendering technologies into the base platform (e.g., perhaps as application-specific integrated circuits), adding simple communications port(s), and designing more capable cartridges or building in a simple DVD player.

Keep It Simple

The notion that computing has to be delivered to people on the other side of the digital divide using PC technology is fundamentally misguided. Using alternate models, like highly integrated but relatively cheap appliances for children or gaming, can kick-start the information revolution in developing countries. Focusing on child-like simplicity, radical hardware integration, and open source software can provide the answer for all who are interested in collaborating to build affordable information processing tools for the people being left behind in the information revolution.