The rapid growth of computing power has given the Air Force many of its pre-eminent capabilities-stealth, sensor fusion, uninhabited aircraft, and precision guidance, to name a few. It has also put the service in a bind. The quickening pace of turnover in the state of the art of digital electronics means that avionics also become obsolete very rapidly, making it more and more difficult to sustain the fleet.

Manufacturers of microprocessors-computer chips-and related hardware used in USAF aircraft are increasingly opting to discontinue production of yesterday's technology in favor of the latest thing. Often, this happens with little warning, leaving Air Force program managers scrambling to either quickly locate an alternate source or hastily throw together an unplanned upgrade to a new standard. Unfortunately, neither fix will guarantee that the service won't have to go through the same drill only a few years down the road. Stockpiling old parts is neither affordable nor desirable, considering the speed with which they become obsolete.

Not even the newest weapon systems are immune to the problem. Managers of the F-22 fighter, still at least five years away from being fielded in squadron strength, must now budget $50 million a year to replace "old" F-22 avionics with new hardware and software. Moreover, the Air Force designed the F-22 to have an open avionics architecture, one that is designed to accept a change out of avionics with relative ease.

Older aircraft pose major problems. For example, the almost 40-year-old B-52H bomber contains a hodgepodge of electronic systems of a wide variety of vintages, including 1950s-era vacuum tubes. The B-52 provides an example of a closed system, one that is unique and unable to operate with other avionics equipment not made specifically for it. The B-52H is not the only aircraft in this situation; 41 percent of the USAF aircraft inventory is more than 24 years old.

Massive Challenge

The problem poses a massive challenge. Lt. Gen. Robert F. Raggio, commander of Aeronautical Systems Center, headquartered at Wright­Patterson AFB, Ohio, gets the job of trying to keep the Air Force on top of the avionics problem. He reports that it has been a snowballing issue for the service.

"Anyone who's bought a computer" can appreciate the problem, Raggio told Air Force Magazine. "We've all experienced this: ... to find our computer out of date a few months after we buy it, [when] there's a new chip that's come out." Multiply that typical household experience by an inventory of more than 6,000 aircraft and the magnitude of the Air Force's predicament becomes more comprehensible. Moreover, all the services are similarly affected. "They're grappling with the same problem we are," said Raggio. "We've got cases where the chips and the piece parts that are in these aircraft literally are reaching the point where we can't even procure them anymore."

One example is a flat-panel display found in modern "glass cockpits." A major supplier of the unit went out of the business last year, affecting a wide array of platforms in several services.

The American defense industry is not trying to be uncooperative, Raggio said. It's just that the business outlook in electronics has undergone a dramatic change. Once, military requirements dominated the demand side of the market in ultra-sophisticated electronics. Now, service needs are dwarfed by the enormous consumer market. The industry follows the market, and the military niche now is simply too small to generate much interest in some companies.

"One day, [computer chip maker] Intel called and said they're not going to make any more government-specific chips," Raggio reported. "Motorola has done that [too], and it's purely a business decision. There's not enough business base for them to [continue to provide the military with unique hardware], so they go out of the business."

In the early 1990s, Pentagon policy and outright necessity pushed all of the services toward the use of commercial-standard parts. In one sense, costs have gone down as a result of using off-the-shelf equipment. It is produced in vast quantities and is unencumbered by reams of documentation and expense attending the creation of a military specification--or milspec--item. In another sense, however, the "cost" has gone up. The services now lie at the mercy of market forces which, like clockwork, make electronic systems obsolete every 18 months.

"Computing power is the coin of the realm" in modern aircraft, according to Gen. Michael E. Ryan, Air Force Chief of Staff. While computing power has been doubling every year and a half, he noted that it often takes at least that long to state a requirement for a new system, get it designed, out on contract, and installed. "That presents you with a problem of being at least one step behind," Ryan said, noting that, every time USAF installs new avionics in a system, "industry out there is already two new issues of hardware and software" beyond.

Cutting Edge No More

A case in point is the B-2 bomber. The stealth aircraft was designed in the early 1980s and is one of the newer platforms in the inventory. Its quad-redundant flight-control system is powered by 286-mHz processors-cutting edge at the time, but now not even fast enough to power computer games for toddlers. This complication affected USAF decisions about how many of the new bombers to procure. Great amounts of money would be needed to create newer avionics architecture with rewritten flight-control software and testing, said budgeteers.

So far, older technology has not yet led to serious shortfalls in combat capability. Gen. John P. Jumper, commander of Air Combat Command, Langley AFB, Va., said he "can't point to one thing ... where we are limited by computational power in what we have either ongoing or planned in upgrades to our systems." However, he added, "there's probably something out there."

Empty parts bins, though, do represent a combat weakness. "It's not a question of having the latest and greatest," said Raggio. "It's a question of whether you're going to be able to sustain [a system] for the next three, four, or five years. It's very short term."

When an update isn't possible, any contractors the service can talk into making the out-of-date parts it requires may charge "an arm and a leg" to provide them, Ryan said. The service can ill afford spending "premium buck" on keeping old warbirds flying with bygone technology, he observed.

Yet, Raggio said, it was just such a pile of mounting, unexpected bills that caused Ryan to direct ASC to confront the avionics turnover issue. The issue got its own acronym: DMS, or Diminishing Manufacturing Sources. The initial marching orders were to develop a plan to preclude the obsolescence of avionics, Raggio said.

"Well, you can't preclude obsolescence," Raggio noted. "That's going to happen whether you plan for it or not. You work with the obsolescence that's going to occur."

Raggio established an Aging Avionics Office, in addition to its Aging Aircraft Program Office. Both fall under Col. Joseph Shearer, who is director of ASC's Subsystems System Program Office. Raggio asserted that the Aging Aircraft Program Office will become, for Air Force Materiel Command and USAF, the "focal point for affordable avionics architectures." A new position of ASC chief avionics architect has been taken on by David G. "Butch" Ardis, who is a technical advisor for avionics systems architecture at ASC. He will work with Shearer and try to pull all of the diverse systems in a common direction.

Toward Open Architectures

After brainstorming the problem, ASC concluded there was only one way to keep current with electronics technology: require that every new electronics system installed in USAF aircraft be of open architecture, meaning that it could easily accept new technology as it arises, and also be capable of working with all related products and of talking to older systems, as well.

Air Force officials compare such a system to the home Personal Computer running with the Windows operating system. The PC can use a wide variety of hardware and software, from a broad array of manufacturers. It can be upgraded and will work with just about any kind of peripheral equipment-scanners, faxes, printers, etc.-because everyone in industry observes common manufacturing standards for cables, interfaces, and wires.

"Everybody's using it, so everybody writes [software] to it," Raggio said. "It also allows you to plug-and-play anything you want to put into it."

Moreover, there is an industry standard for connections, impedance, and other factors that make all the gear work together well. All of these things-interoperability, a diversity of vendors, potential for growth, and adoption of a recognized standard-will be incorporated into a new program directive from the Air Force leadership, mandating the use of open avionics architectures on every flying platform.

The Air Force argues that requiring open architectures in avionics will lower costs by expanding the number of vendors that can compete for work as well as reduce life-cycle costs by using common equipment rather than one-of-a-kind types.

Plans called for Ryan and Air Force Secretary F. Whitten Peters this spring to sign and release a letter outlining the policy to program managers. A high-level directive was required, because program managers have to devote resources--personnel, as well as financial--to make open architecture work, and those resources will have to be pulled from other assigned tasks.

"I needed advocacy for this initiative," Raggio said.

While some programs have, on their own, already made great progress in open architecture, some have devoted very little study to it. "Now, you might say, why didn't we do this all along?" Raggio asked rhetorically. Answering his own question, he observed, "We have gotten a lot smarter about what constitutes an open system in the last few years. We thought we understood open systems back in the 1990s, but we didn't really." Computer architectures are "a moving train." They change very rapidly, he said, noting a successful open architecture policy "was probably not possible a decade ago."

It's one thing to require that future avionics be open, but to make it really work, there has to be a plan to make it happen. Ardis told Air Force Magazine that each aircraft program office will have to produce a specific plan for reaching an open architecture by mid-2001. These subplans will add up to a single master plan, which will be a major factor in calculating future budgets, considering the massive dependence of the service on electronics.

One Part at a Time

Raggio said that, given the constraints on the Air Force budget for the foreseeable future, it's doubtful that the avionics problem will ever be fixed totally. Instead, each aircraft will have to be brought up to par incrementally, in an "evolutionary" way, until each is open to easy avionics upgrade and replacement. This will require fixing one part at a time--a radar, perhaps, or a diagnostic system.

Ardis noted that the F-15 is getting a new radar. It won't solve the vendor problems with the F-15 but will eliminate much of the headache and cost of fixing the airplane when it breaks. Savings can then be devoted to other upgrades. The radar "is a totally open system embedded in a closed system," Raggio said, making it an open architecture subsystem. As these continue to be added, the whole system becomes "more and more open."

Ardis pointed out that "there's tremendous benefit" to even a partial improvement. "If we could take the items that are driving our support costs the highest and work those off first, ... there will still be a tremendous benefit to us," said Ardis.

The C-17 airlifter is the beneficiary of an interim step called "software wrapping," a technique in which a closed avionics system "can appear to other systems as open," Raggio said. This alternative method is being explored in Air Force labs. The objective is to get to a fully open architecture as soon as possible.

Open architecture allows not only for cost reductions but for adding new capabilities quickly, as they become available. These updates will have to happen more frequently than they now do if USAF is to maintain pace with the state of the art.

Raggio said the new open avionics strategy will likely propose a two-year turnover in avionics on those platforms that are most in need of highest technology--most particularly, fighters and bombers. He likened this to the current "tape change" in which, every year or so, new threats, capabilities, weapons, and improvements are added to the software of combat systems. In addition to software, Raggio envisions a hardware change.

Today, "we have modernization requirements coming in at various times" after the initial development of a weapon system, he observed. Some of these "blocks" are large, some small, "and they're really not timed to any point in time. They're usually when you can afford it."

An upgrade could stem from an operational shortfall against a new threat or an opportunity to drastically reduce costs by shifting to a newer, more easily maintained system.

Threshold of Pain

Under the current acquisition scheme, upgrades are budgeted as funds allow. However, when a vendor suddenly departs the business and leaves no supplier or airplanes don't have a desired readiness rate-"when the threshold of pain gets high enough," as Ardis puts it-operational forces come demanding an upgrade. Because the change is unplanned, the opportunity for a well-thought-out, cost-saving, and effectiveness--boosting upgrade is sidelined by a mad dash to get the airplane back up to its status quo.

"The Air Force, I believe, has got to make the commitment ... to upgrade the avionics of weapon systems on a periodic ... block basis, at a scheduled time, ... and that block upgrade will include hardware and software enhancements," Raggio asserted. Vendor problems can be assessed at each upgrade and solved as they occur, "but only if we do this as a conscious effort and as a conscious plan." This, he said, "is the only way we can figure out to get ahead of and stay up with the avionics changing challenge."

He acknowledges that this would be ambitious and may require a bigger chunk of the Air Force budget than is now available for avionics improvements.

Commanders in chief and the heads of Air Combat Command and Air Mobility Command would have to make the "operational trades" between competing priorities for the spending of money, he asserted. Any upgrade deferred would go "into the pot" for the next round, but there would have to be an understanding that "delaying something to another block has consequences," since this would only push the system further behind and create more opportunities for a parts crisis in that system later.

The Air Force has sponsored two no-cost studies, one led by Boeing and one by Lockheed Martin, on the feasibility of moving rapidly toward an open avionics architecture. The Boeing effort, called the Open Avionics Systems Integration Study, looked at Boeing platforms including the B-1, B-52, C-17, F-15, and, with Northrop Grumman, the B-2. While results are not yet in, "we do know that the more common" the avionics systems of the fleet are, "the better this solution is," Raggio said.

At Lockheed Martin, a study called Systems, Technologies, Architectures, and Acquisition Reform is also examining the possibilities of open, common avionics in the F-16 and F-22. This study is two-thirds complete, but the company reported it might well be able to go beyond the two fighter types, Raggio said.

Formalizing the idea of routine, by-the-calendar hardware upgrades will require rethinking a sacred cow of program management--the Operational Requirements Document.

The Air Force will have to learn how to accept "ORDs that are structured to allow for incremental delivery of capabilities," Raggio asserted. No longer can a new program requirement be stated in terms of the ultimate capability wanted. Programs take too long and the technology will turn over too often before completion for that approach to work anymore. New systems will forever be a hybrid of many generations of electronics, updated at a consistent pace. If the finished, all-up system is demanded at the outset, the program will collapse.

"The first time we test, ... if the ORD says 'full capability,' the first increment fails the test," Raggio asserted.

The early managers of the F-22 program did not structure it along the lines of incremental capabilities, Raggio said, but the new Joint Strike Fighter is being managed that way.

"You can see that we're coming around to this, but this is a real mindset change," Raggio said. The original threshold capabilities for the JSF won't be reached until the third version of the system. "The first couple of items won't be full up," Raggio noted.

Not Affordable

Industry agrees with ASC that "the status quo will not be affordable," Raggio said.

From primes down to the lowest-tier subcontractors, the consensus has been that an open avionics architecture serves everyone's purpose: It keeps the aircraft up-to-date, and it also creates a steady, predictable cash flow for the companies doing the work. The predictability of the upgrades will allow for better preparation of bids, more cost-effective contracts, and lower costs.

"There is a definite vested self-interest" in companies buying into a more open avionics architecture, Raggio pointed out.

"The easier it is to upgrade systems, the more readily we will be able to spend the money to upgrade them. If we can't afford to upgrade them, they [contractors] don't get business."

The acquisition system plays a major role in the move toward open avionics architectures. Raggio reported that the request for proposals on a new, very large-scale C-130 avionics update was snatched back at the last minute because it didn't address the open-systems goal. It was revised and released with the new criteria. "It was a tactical win for us," he asserted. "That is what we'd like to do on all systems."

Ardis and Shearer are working with the various system program offices to create templates for new competitions that mandate an open avionics architecture in all future buys. The language of solicitations has to be just right, and there will be more attention on making sure industry understands exactly what is wanted, Raggio explained.

Fixing the avionics problem will require a master roadmap that will vary with each system in it, Raggio said. The Joint Strike fighter--still a relatively clean sheet of paper--provides the opportunity to start with an open avionics architecture, thus setting the upgrade rhythm right from the start. Older aircraft, however, will take a long time to bring up to date.

The B-52, with "hundreds of [line-replaceable units] ... [will] take a lot of effort over a protracted period of time to get it in the projected funding profile," Ardis observed.

Raggio said the directive from Ryan and Peters will set many things in motion.

System program offices and major commands have "to start planning on these future program increments and to start budgeting" for the upgrades. "[The upgrades] have to fight their way through the [budget] deliberations just like anything else does," Raggio said. "This whole thing has to become part of the formal acquisition program baseline process."

The integrated roadmaps for all systems will force budgeteers to come to terms with the cost of avionics over the whole life of the system, not just answer a short-term operational requirement.

"These kinds of things don't happen unless you make them part of the requirements to be briefed at program reviews," Raggio said, and he has recommended that the avionics issue be part of the quarterly briefs Ryan and Peters take on every major system.

Raggio said the old way of doing business simply won't work anymore.

"The time constant has shrunk," he observed. Technology is advancing at an unprecedented rate.

"When we had the F-4 [Phantom], and a new avionics system came out, we would just do a life-of-type buy" of spare parts, Raggio explained. "We'd say, 'That's good enough. ... That'll last as long as we're going to have the F-4. Another airplane will come along in a few years' " and capture the new level of technology.

However, he went on, "That isn't the case anymore. ... A new airplane isn't coming along" as often as in the past, and the new technology is arriving "at a much more increased rate."

You have to engage and fight the obsolescence battle, Raggio asserted. "You don't have any choice."