Profile: Gary Stupian

What Happened?

by Donna J. Born

A little detective work—and a lot of advanced technology—helps Gary Stupian trace the root causes of electronic system failure.

Gary Stupian, a senior scientist in the Microelectronics Technology Department, came to The Aerospace Corporation in September 1969 after completing two years as a postdoctoral researcher at Cornell University. For Stupian, Aerospace offered opportunities to continue his scientific research in a variety of technical areas. "Aerospace has always stressed research, even in lean times," he said. "It's a very diverse environment, which is what makes it a very interesting place to be."

Gary Stupian, Research Scientist.

He has worked on many programs during his years at Aerospace, but since the mid-1980s, his focus has been root cause analysis, the systematic investigation into a problem or an anomaly to find the underlying physical cause in order to fix it and prevent its recurrence. Stupian said such analysis has historically been part of the corporation's work in maintaining currency in space technology. "Root cause analysis covers all programs. Like an undertaker in a small town, we eventually get everyone's business."

One of the company's leading authorities in this area, Stupian described this work as technically challenging, eventually involving "the application of essentially every scientific discipline that one has studied." Space programs do have failures, and most of the technical staff inevitably will spend much of their time trying to understand and correct them, he said. "They will get down to the atomic scale to find out what's going on."

"That's what we do, myself and other people in the labs," he explained. "We generally drive for the absolute 'for sure' cause of why something didn't work, why it failed. You can often do that, but sometimes you have to be satisfied with the probable cause. You don't want unverified failures—everyone lives in fear of an unverified failure. You don't know whether it's going to come back again. You'd like to know what really did happen, and then you can either work around it or take corrective action."

Investigating the cause of anomalies is tied to the very beginnings of the corporation, when one of its earliest assignments was to assume system engineering for the Atlas launch vehicle and improve its reliability from 85 to 99.9 percent to make it safe for human flight. The Atlas, destined to carry the first astronaut into space, had already failed twice, once just a week before the corporation was formed in June 1960. Comprehensive design analyses led to modifications that improved the reliability of the launch vehicle, and the Atlas successfully lifted John Glenn into his historic three-orbit flight aboard the Mercury capsule in February 1962.

Designers try to catch failures during testing before launch, Stupian said: "In space nothing is really reparable." Testing does cost money, and tests have to be designed with great care, he cautioned. The number of parts that can be tested under temperature and vacuum, for example, are limited, and testing has to be "accelerated," most commonly with elevated temperatures, so results are timely enough to be useful. The spacecraft is made up of components, and each one has to be reliable. Catching problems at the component level is the least expensive solution. Fixing problems becomes progressively more expensive at higher levels of integration and as launch dates approach, he explained.

Failures revealed by testing are examined carefully using advanced laboratory techniques. Stupian said that analysts will ask questions such as: Are the failures in an accelerated test representative of real, end-of-life failures that will limit a mission, or did inappropriate testing break the parts in some other way that won't be a problem in the application? If a failure involves a component installed in hardware, maybe even on the launchpad, what went wrong? Is there a generic problem that will affect all similar parts, or can the failure be attributed to mishandling that is not likely to be a recurring difficulty?

"You hope you have only one failure, but you must know the root cause if mission success is to be guaranteed," he said. "Sometimes, failures result from some very familiar physical mechanism; in other cases, the failure may result from a process that hasn't previously been responsible for anomalies. The ones that are not totally routine are more interesting, as a rule, but you have to look at everything."

In recent years, root cause analysis has acquired even greater importance, Stupian said, because of the decline of military influence on the electronics industry, reduced funding, and to some extent, offshore fabrication. Military requirements used to drive the supplier industries, especially the semiconductor market, but with the growth of consumer and industrial electronics, the small aerospace industry has little clout to dictate what suppliers are willing to provide. Manufacturers can change designs anytime—a component that previously worked may no longer do so; a change to facilitate production may be disastrous for military space but inconsequential for consumer applications.

"I think in the present age, when we are switching emphasis to commercial parts, Aerospace is needed more than ever. We must look at the roots of failures to help understand how to make these commercial parts viable to ensure the success of space missions."

Stupian predicts that smaller parts will bring further challenges to root cause analysis: "Semiconductor feature sizes are shrinking. We're at about 0.25 micrometers now. By way of comparison, a human hair is about 75 micrometers in diameter. You can stack 300 such tiny objects (e.g., transistors) side by side and just span a single hair. Devices with 90- and 45-nanometer feature sizes will be used in systems now being built. We're able to do some rather remarkable things, including complete 3‑Dimensional reconstruction and modeling of nanoscale structures and chemical analysis on the nanometer scale. This sort of challenging work will grow in importance."

His work in the area of reliability and root cause of reliability problems earned him the Aerospace President's Distinguished Achievement Award in 1994. His expertise is regularly in demand, and he has been involved in numerous investigations. For example, hybrid circuits in the Milstar flight computer were replaced based on evidence he collected working with Tom Hoskinson of Aerospace's Milsatcom Division. His work with microfocus radiography (also called X-ray microscopy), which provides real-time imaging of details of the internal structures of specimens, is well known in the contractor community.

Stupian has been with Laboratory Operations during his entire career at Aerospace (and has kept the same metal desk through several organizational changes), where in addition to his work with root cause analysis he has been involved in many aspects of surface science, "including Auger spectroscopy and scanning tunneling microscopy." He regularly publishes in scientific journals, including articles this year on fabricating a photonic crystal and on high-pressure physics.

He considers some of his most interesting, "albeit rather tangential," work to be in forensic science. He has assisted investigations with the California Highway Patrol and the Los Angeles Police Department. In one murder case, he and Neil Ives, also of the Microelectronics Technology Department, worked with the coroners' investigators using X-ray computed tomography ("similar to a medical scan where you take a cross-sectional view and then you can stack the sliced images to form a complete 3‑Dimensional model") to examine the vertebra of a murder victim.

Stupian discusses a forensic investigation with members of the local police force.

He was the first to look at the isotopic composition of bullet lead to characterize bullets, and published papers on the subject in 1975 and 2004. The nuclei of lead atoms can have different numbers of neutrons; that is, there are different isotopes, he explained. The four main stable isotopes of lead are found in varying relative amounts in nature because of differences in the initial chemical compositions of the precursor radioactive minerals. Lead from different geologic sources will show differences in the isotopic ratios.

"This has become controversial now," Stupian said. "Some laboratories have made very strong assertions about their ability to do this type of analysis. They were looking also at variations in elemental composition, but the same principle applies. The trouble is that the lead used in bullets may be recycled, so it's very hard to vouch for its uniformity."

From the time he was 10 years old, he wanted to be a physicist. Although his long working hours leave him with little free time, he spends much of it keeping up with developments in physics outside his area of concentration—for instance, dark matter and dark energy in the universe—"because a physicist ought to know these things." His three academic degrees are in physics (with specialization in condensed-matter physics): B.S. from California Institute of Technology and M.S. and Ph.D. from the University of Illinois at Urbana/Champaign.

He has been active in helping young scientists and recruiting them to Aerospace through his work with the corporation's university affiliates program, which promotes the exchange of technical information, expertise, and research with selected universities. As the technical liaison between Aerospace and Caltech, he is largely responsible for obtaining funding for six undergraduate research fellowships each summer. The students work on the Caltech campus with faculty, graduate students, and postdoctoral fellows over a 10-week period during the summer. The six company-sponsored students are asked to present their work in seminars at Aerospace at the end of the summer. They get to see the Aerospace campus and gain a broader awareness of the company's role in national security space. The Aerospace Institute, the division of the corporation that administers the university affiliates program, has recognized his significant contributions to the program with four annual achievement awards.

"Aerospace is a good place for young scientists to pursue a career," Stupian believes, "because it is one of the few places where you actually have the possibility to do research. We're involved in the practical things, and we try to do some more fundamental things. There are not very many places where you can do that. We don't do as much research as we would probably like, but I have a lot of satisfaction in helping space programs by applying physics and material science to solving problems."

To Fall 2005 Table of Contents