Pervasive Prognostication

Shortened, cold days in some parts of the world only compound the dreariness and gloom under the specter of terrorism, layoffs, and first-quarter reports. What a good time to take out my crystal ball and provide some positive pervasive prognostication.

Gazing into its glassy depths, how is it that I feel as if I’m looking backwards, not forwards? With the OMB, the Fed, and others finally admitting that the United States is in, gasp, RECESSION, new Middle Eastern conflicts, and a Bush in the White House, it feels like 1991 all over again. And that’s great news for those who are in the business of embedded and pervasive computing.

Under recession in the early nineties, embedded tools and platform vendors actually prospered. Companies like Microtec, ISI, and the nascent Wind River had 40% or higher annual growth rates—them were the good old days. Multiple factors conspired to make this so.

First, one side-effect of the downsizing that comes with recession is the surviving company’s need to optimize and to outsource to replace capacity they’ve let go. Before the downturns, companies were (and more recently) eager to invest organically in their entire value chain and in vertical hardware and software stacks. Tightening their belts, however, they become keenly aware of their true value add and they revisited previous build-vs-buy decisions. In pervasive computing, this refocus means that, rather than develop their own platforms, protocol stacks, middleware, and applications, they begin to look to outside vendors. What they keep is their unique intellectual property—hardware technology, manufacturing capabilities, and high-value software content.

These trends have a mixed effect on embedded platform and tools vendors. In fat times, they could close deals quickly and sell plentiful development seats. In leaner times, they must provide more stackware and accompanying professional services. Existing customers end up looking more like integrators than developers, resulting in average sales that address fewer developer seats. New projects do crop up as engineers who are let go from one company show up at another or start efforts on their own.

Second, at least half of embedded vendors’ key customers will continue to invest in an R&D project, building for the “next wave.” These projects may start slower and take longer to deploy, but when recovery comes, vendors who have stood by their customers always benefit.

Third, unlike many enterprise-oriented business sectors, embedded/pervasive companies have the agility to quickly refocus their marketing efforts to new segments. For the last eight years, big telecom and wireless were the darlings of investors and easy pickings for embedded solutions companies.

Lucent, Ericsson, Alcatel, Nokia, Cisco, AT&T and others were starting new projects almost every week and the embedded crowd couldn’t ship tools and kernels fast enough. Now, with telecom stars in the doldrums, embedded vendors are focusing on next-generation consumer devices such as handheld/wireless, set-top boxes, and automotive telematics, which are also design-ripe for Linux.

Another area for refocus is aerospace and defense. After a decade-long hiatus, new projects such as the controversial ABM, the Joint Tactical Fighter, and smarter munitions are reopening doors at primes and subs with multiple spinoff and support opportunities.

Lastly, in all areas of embedded and pervasive computing, developers and management are today looking to streamline. They seek cost-savings in Open Source in general and in embedded Linux in particular. This sheltering in Open Source, however, is a mixed blessing. With greater competition for scarcer projects, companies like Microsoft have embedded Linux squarely in their cross hairs. The powers that be are worried and have declared open season on penguins.

Luckily for embedded Linux, they’re still shooting with 10-year old guns.