“Modified COTS” is something of an oxymoron . . . but we are unquestionably seeing the beginning of a trend that, if unchecked, could set the military procurement market back by 10 years.
By Peter Cavill
In 1994 when then-Secretary of Defense William Perry formalized the requirement for defense suppliers to use commercial off-the-shelf (COTS) solutions, his vision of the benefits he foresaw was clear. Development costs would be lower; equipment would cost less as a result of moving away from mil specs and of competition between suppliers; and expenses would shrink as technology insertion mitigated the effects of obsolescence, or COTS suppliers absorbed the costs.
Perry doubtless also believed that in setting a standard for the United States he was implicitly setting a standard for the world. The U.S., where spending on defense currently constitutes 46 percent of all defense spending worldwide, is by far the world’s largest procurer of defense solutions.
Ten years or so later, however, is Perry’s vision a reality? Is there international consensus on the advantages from a procurement strategy that mandates COTS solutions? Are the products that are being procured truly “commercial off-the-shelf”? And is the competition that Perry foresaw as a key driver in reducing costs a reality?
COTS: an international reality?
In the United States, the end customer most often mandates prospective suppliers to tender COTS solutions, especially in rugged applications, which makes the United States a relatively straightforward market for a COTS vendor to operate in. Complex requests for quotation, with specification down to the component level, are now relatively rare, and systems integrators will accept COTS specifications and standard qualification reports.
The effect of competition has worked well in the U.S. COTS market; increasing numbers of systems are built using several different COTS vendors. In-house design is a position of last resort where specific functionality is unobtainable via any other route. Technology insertion has proven itself to be a substantial COTS benefit; several examples of programs were developed on one generation of processor, qualified on a successive generation, and deployed on a third. Radstone Technology in Towcester, England, for example, is marketing the fourth generation of a backwards-compatible processor board originally introduced in 1994.
Europe, however, has a different position. Even though this region is the world’s second largest spender on defense, its combined total spending of thirteen countries is well under half the spending of the United States. In Europe, COTS solutions are recognized, but are not necessarily preferred. Standard interfaces like VME continue to be widely specified but COTS continues to be only a small part of the whole. Whereas a system developed by a North American contractor can be largely based on the contribution of a number of COTS vendors, a European contractor will likely design the majority of the system in-house with perhaps only a single function being performed by a COTS component. Custom design based on a COTS form factor is a common requirement, but smaller program sizes can make this uneconomic for a COTS supplier.
European end customers often demand the impossible-or at least, the impossibly expensive. A requirement for all ceramic mil-spec components, for example, is not rare. Even when COTS components are selected, European customers will often wish to approve the design. Where technology insertion has been widely embraced in the United States, moreover, European customers express much more concern over the ongoing provision of identical replacement products.
Asia is, today, a relatively small spender on defense: the combined spend of China and Japan, for example, is around 25 percent of the spending for 2004 in the United States. It could be that this is why the Asian market has embraced the COTS model so completely-perhaps more so than the United States. In Asia, the typical requirement is for a “one-stop-shop,” turnkey, application-ready solution built around COTS boards and with the chosen supplier required to procure, fit, and test the necessary components. Decisions on technology insertion are devolved to the supplier, who is expected to provide appropriate long-term maintenance and support.
So is it only Europe that is out of step with the rest of the world? In fact, the situation is changing. There is a growing acceptance of the benefits of COTS, and there is a growing realization that COTS solutions are the only ones that are economically viable. Trust is growing in the ability of COTS suppliers to provide appropriate solutions. In this one respect at least, the United States and Europe are converging. After an extended period in which the world’s two largest spenders on defense were pursuing different procurement philosophies, the COTS approach appears to be becoming a worldwide reality.
COTS: a commercial reality?
As this convergence in philosophies is occurring, however, are we seeing the beginning of a divergence away from the original principle of COTS? The original vision of commercial off-the-shelf solutions involved a set of components that are widely available across a broad range of applications and in a broad range of markets. Standard architectures and components would mean maximum compatibility, easy interoperability, cost-effective upgrade, and the widest range of continuing options. Eliminating application-specific solutions would reduce development, deployment, and support costs.
Surely, then, “modified COTS” is something of an oxymoron. By definition, once something is modified, it is no longer standard, and the benefits that accrue from adherence to standards are lost. This is an overstatement, of course, but we are unquestionably seeing the beginning of a trend that, if unchecked, could set the military procurement market back by 10 years. The requirement is increasing for solutions that are, superficially at least, COTS-based-but with program-specific modifications to functionality or input/output. To this point, more than 50 percent of Radstone’s business now derives from customer requirements for modified COTS.
The competitive nature of the marketplace has, until now, been one of the causes of this. The question goes something like this: “If Vendor A has these attractive features, and Vendor B has these other attractive features-why can’t I get a product that combines both?” The financial muscle of the major defense contractors has left COTS vendors with little room to maneuver, and two options. Option one is the vendor must invest significant additional resource in ensuring that his product range is as broad as possible, and reflect as many of the contractors’ “care-abouts” as possible. The second is the vendor must put in place the resource and infrastructure to bid for and execute modified COTS designs.
Those challenges are not becoming easier. An additional driver in the market is the accelerating development of new technologies that, superficially at least, offer benefits that cannot be ignored. If there has historically been stability in the defense electronics market-and thus, by extension, the COTS market-it has derived from the adherence to standards such as VME and Ethernet.
The emergence of technologies such as StarFabric, Infiniband, RapidIO, and PCI Express is giving the whole industry pause for thought. Each has significant attractions-but also weaknesses. Will any one or two of them-or all of them-emerge as industry standards? VME has survived because it has demonstrated a Darwinian aptitude for adapting to its environment.
VITA-the VMEbus International Trade Organization-has been quick to evolve responses that have accommodated new technologies, and continues to do so in the field of high-speed fabrics and system input/output. Peaceful coexistence looks a likely outcome-but in the interim, COTS vendors are placed in an unenviable position as they try to ascertain the best place to apply scarce resources in response to diverging customer demand.
The competitive market raises customer expectations and new technologies are introduced with gathering speed. The result is that what were once- standard products are becoming less so. While the industry today is far from where it once was in terms of reliance on proprietary solutions, it has moved subtly-but identifiably-one or two steps backward. It is a trend that would, if unchecked, prove to be counter-productive. The reality seems, at present, to be that there will be little further movement in this direction.
COTS: a competitive reality?
Geographically, the COTS market has diverged, although it looks as if it will converge. Commercially, the COTS market has diverged from William Perry’s vision-but today the divergence is fractional and the work of groups such as VITA is likely to ensure that it remains so. What, then, of the competitive state of the market?
Here too, divergence is apparent. On the one hand, vendors of new technologies are coming into the market and changing the competitive landscape-unquestionably, for the better, even though the short-term implications for standards are not yet clear. The more suppliers in the market, the more competitive it will become. The winner, of course, is the customer.
But there is a distinct and contrasting trend in the market that should worry customers-the trend towards consolidation, towards acquisition, towards reducing rather than increasing the number of players.
Curtiss-Wright completed its sixth acquisition; Dy 4 Systems joined Vista Controls, Systran, Synergy, Primagraphics, and Peritek in the Curtiss-Wright stable. Like a small boy with baseball cards, Curtiss-Wright seems to be trying to complete a full set. SBS was created through the acquisition of 11 small companies. GE-Fanuc has entered the market by acquiring VMIC, Ramix, and Computer Dynamics. Motorola bought Force (although likely to increase Motorola’s telecommunications capability). Even Radstone is not immune from “acquisition syndrome,” having taken ICS and Octec under the corporate wing. In those few examples alone, close to 25 potential suppliers to the defense market have been reduced to no more than five.
Superficially, with so few suppliers from whom to choose, the outlook for customers has deteriorated. The reality, however, may not be quite so bad. It seems probable that these new conglomerates’ aspirations are further up the supply food chain (where they will likely compete with companies who were previously their customers)-and by so doing, they will create opportunities in which new entrants to the market can not only survive, but thrive based on their three competitive advantages: their innovation, their nimbleness and their aggression.
In summary, the prognosis for COTS solutions in the defense market is positive. Where potentially damaging geographical divergences existed, a new unity of vision is becoming apparent. Divergence from the classic definition of “commercial off-the-shelf” is perhaps causing some short-term dislocation-but it is helping to create a vibrant and innovative market. And the divergence in the competitive environment-with the creation of “supergroups” on the one hand and the emergence of new companies leading the market in emerging technologies on the other, ultimately benefiting the customer. And any market that is consistently delivering benefits to its customers is, by definition, a healthy one.
Peter Cavill is managing director of embedded computing specialist Radstone Technology in Towcester, England.
