Bridging Hardware and Software

From an interaction designer's point of view, the divisions between hardware and software are inconsequential because they are inconsequential to a user. The user doesn't care which is more expensive to build. Thus, interaction designers can resolve problems that arise during development of hybrid products.

In the world of engineering, there are hardware engineers who create circuit boards and microchips, and there are software engineers who create program code. Although the fruits of their labor are sold in a common or hybrid product, the two factions typically don't work cooperatively. Sometimes they don't even communicate, but instead merely throw completed modules over the fence into the other's backyard.

For historical reasons, hardware engineers dominate most hybrid-product companies, but as the availability of hardware increases to the point of ubiquity, hardware and its engineers assume a less critical role. Conversely, the true value to the user of most products is increasingly due to the unique contributions of the software. This makes for an uneasy truce in most hybrid-product companies.

Hewlett-Packard is a good example of a hybrid-product company dominated by hardware engineers. Its printers are fabulous products with exemplary engineering, but after two decades of refinement, none of my HP printers yet cooperates fully with my computer. They don't bother to tell my computer how much paper is in their feed bins, how much ink is left in their cartridges, or how many print jobs are currently queued up and ready to go. This kind of thoughtless disdain for the human's need for information is the telltale smoking gun of companies dominated by hardware engineers.

Ironically, hardware companies are more experienced at seeking outside help from industrial-design firms to help them make their products more desirable and useful to their users. Software companies tend to go it alone. In any company making hybrid products, when hardware and software engineers don't have designers to mediate between them, the result will be products that fail to please. Most of the examples in Chapter 1, "Riddles for the Information Age," make this clear.

As more and more products are hybrids of hardware and software, the need for a Goal-Directed design increases, because it is agnostic regarding the implementation medium.

3Com Corporation original maker of the PalmPilot is a good example of a hybrid-product company where design created a smooth integration between hardware and software. A single tap of the screen and the machine awakens instantly in the exact state it was in when it was last shut down. When hardware is instantaneously responsive to users' actions, it is a clear indicator that the hardware design incorporated the needs of the software. Conversely, my Nikon CoolPix 900 takes seven long seconds to boot up every time I turn it on, and it doesn't even have a disk drive. When hardware is this sluggish, it is clear that the hardware engineers ran the show.

Of course, in the real world of product design, most software companies quite rightly stay out of the hardware world. Designers respect this, even when dedicated hardware would confer significant advantage.

However, if the cost structure of the design problem allows for it, designers should not hesitate to make recommendations about hardware. The Sony P@ssport IFE system in Chapter 9, "Designing for Pleasure," ran on dedicated computers, and the vendor had complete control over all hardware and all software. My designers made several hardware recommendations.

In the Elemental Drumbeat design in Chapter 10, "Designing for Power," the product was destined to run on any vanilla, Wintel desktop computer. My designers stayed well away from any hardware recommendations.

For several client projects, including Logitech's Peacock, my designers were lucky enough to find opportunities to add value with hardware design. Each company had the option of venturing into the world of hybrid solutions, with all of the danger and opportunity that that entails.