Related: Digital Domain
It was more than 15 years ago when the U.S. Department of Defense began toying with the idea of developing a single fighter jet platform to fulfill the needs of the various wings of the military as the luster of the F-18 and F-22 began to fade. As studies began, the military quickly discovered several allies—most notably Britain—interested in developing their own new fighters, which begged the question: Why not develop a single fighter platform that can be shared with various allies to meet their performance requirements? The benefits were pretty easy to identify: shared risk in development costs would be borne by all partners, increased access to suppliers worldwide, and reduced overall per unit procurement costs. Thus started the development of what is now known as the F-35 Lightning II, commonly referred to as the "Joint Strike Fighter" (JSF). This project would impact more than the halls of the Pentagon or the British Ministry of Defense, it would change the way businesses across the globe manage their plants and product development teams.
On October 26, 2001, the U.S. government pulled the trigger on the F-35, awarding a 10-year development contract to Lockheed Martin to bring the JSF to reality. With the Cold War over and the U.S. government paying close attention to its coffers, JSF expanded its reach to encompass other allies beyond the British—including Italy, the Netherlands, Turkey, Canada, Australia, Denmark and Norway—who signed up to get the latest in fighter technology at what would be a somewhat affordable cost—approximately $45 million to $60 million—compared to the outgoing F-22's $99.7 million per unit price tag . While the per unit savings may have been more than enough to make even the most fiscally conservative members of Congress gleeful, it meant a new series of obstacles for Lockheed, since the new partners would need access to top-secret data, not to mention the necessity to spread the wealth in terms of supplier contracts within the partner countries. A key challenge, however, was developing a computer program that could handle the complex data and modeling systems needed to bring a machine with 1.5 million parts and 3.5 terabytes of corresponding data per unit to fruition, while complying with the U.S. government's International Traffic in Arms Regulation (ITAR)—a law that provides the President with the sole discretion on the exportation of defense-related articles and services. ITAR required Lockheed to limit shared data to only those areas which the partners are involved: "Certain partners are allowed to see only what they need to see to design their portion of the aircraft. The classified components are only seen by those who are cleared to see those components of the aircraft, but the structure itself can be seen by everyone," says Joe Fowler, director of PDM implementation for Lockheed Martin.
At the time the JSF took root, Lockheed needed to seek out a partner who could manage the complexity of the most ambitious global defense system project ever attempted, giving the nod to UGS Corp. (Plano, TX) to supply the backbone used to develop the F-35, what the aerospace firm called its "digital thread." This thread would have to be used to validate designs for every major component, eventually encompassing numerous languages to connect the more than 1,000 suppliers and 20,000 users in more than 30 countries 24 hours-a-day/7 days-a-week. It was a task that was even foreign to UGS at the time, but Lockheed had already begun developing the software code that would make what's known today as "product lifecycle management" (PLM) a reality. "When we started this thing, Teamcenter software was just in the meta phase," says Fowler, who quips the reason why the acronym PDM (Product Data Management) remains in his title instead of PLM is because that when they started, PLM was not in the vernacular, and he figured that making the switch from PDM might signify that something was wrong with the original concept, which is the furthest thing from the truth.
Designing PLM software to support such a complex product that is expected to have a 50-year lifecycle has turned out to be a valuable learning opportunity for both UGS and Lockheed. "The sheer size and volume of the data we had to manage was immense and we had to continue to stay ahead of them at every step of the process," says Tim Nichols, who leads the aerospace and defense global center of excellence at UGS. "You have to remember, this is an innovation they have never used before, so we had to get them to really buy into it and they saw the benefits early on, but there was a bit of resistance to change." The results through the initial phase of the program have been pleasantly surprising as the merging of the software and physical structures of the airplane in the virtual world has helped Lockheed and its supply base identify additional savings of $62 million throughout the development process, while improving parts sharing among the three JSF variants—one designed for conventional takeoff and landing, another for aircraft carrier and shorter landing scenarios (which require heavier airframes and a larger wing area to meet the demands of catapult launch and cable landings) and another designed for vertical takeoff and landing—to as much as 80%, along with a 35% reduction in cycle time during the design phase, providing billions in additional savings.
With the design of the JSF basically complete, the team must now focus on building 15 prototype aircraft before the production units are delivered in 2009. Prototype production has just been completed on the first plane—a conventional takeoff and landing unit—with flight testing slated to begin in November. The move from design to manufacturing is expected to be equally as challenging for Lockheed and UGS, as the major body pieces for the JSF will be constructed in three facilities (not to mention the hundreds of suppliers that will provide smaller pieces, including aeronautic control systems, wiring harnesses and control mechanisms): Center fuselage is subassmebled at Northrup Grumann's El Segundo, California, facility, while the aft fuselage and tail will be subassembled by BAE Systems in Samlesbury, England, and Lockheed Martin will build the forward fuselage and wings—of which the upper section is comprised of a single piece made from composite materials—at its Fort Worth, Texas, facility, where final assembly also will take place. Another final assembly line be added in the future in Cameri, Italy, to build JSFs for European allies, excluding Britain. The final assembly process—of which 80% was devised through digital mockups—will be done via dynamic fixture processes using reconfigurable jigs for each of the variants. The use of flexible manufacturing and subassemblies allows Lockheed to cut final assembly time by 66% compared with the F-22. Complete assembly of the subsystems has been cut from 15 months to 5 months. All of these milestones were accomplished through the use of PLM systems. "As we add manufacturing it requires us to share increased volumes of data," says Lockheed's Fowler, who acknowledges production analysis is placing "pressure" on the PLM system, especially since the myriad of partners requires the system to be up and running 24/7. "We needed this round-the-clock capability from day one, so we focused on the response time of the system, which as to be reliable or else no one would use it," Fowler says. Scalability will become an ever increasing issue thanks to the extended life JSF and the multiple part changes that are expected to develop along the way as technology evolves. Each of the major parts of the aircraft will receive their own serial numbers, which will be complied in a data package that will be provided to the purchaser once the aircraft is delivered. "We now have to manage the build of materials in the as-maintained aftermarket environment while at the same time parts are changing at a pace of upwards of 100 per day and we have to maintain all of that. That means we may have to pulse the supply chain so that they know that in July, let's say, you will need to stop delivering this part configuration and start delivering this new product. That's also why the system architecture has to be flexible enough to evolve with technology to accept various JT, PML and XML files," says UGS's Nichols.
Along every step of the way, the breakthroughs made with PLM through the JSF program have migrated down to benefit various manufacturers. "The auto industry is adopting the approach towards global sourcing, manufacturing and collaboration, just as we have seen in aerospace and defense," Nichols says. These partnerships will require more open communication, with protections for intellectual property to the individual companies. Likewise, as more and more embedded processors and other programming technologies make their way into automobiles, manufacturers will have to track these systems to maintain their compatibility with future technologies, including new telematic systems. Lockheed's Fowler doubts manufacturers would be able to meet the challenges of globalization without the capability of PLM, which he's seen grow from a simple data transfer system, to one that manages manufacturing and design complexities that would have been out of reach a few years ago: "Everything we have done with PLM and JSF is applicable to other industries that have complex products. I know we couldn't have even begun to tackle this project without this capability."