September 18, 2000

A CAREER Award is one of the most prestigious credentials a young faculty member can earn, reflecting the emphasis the National Science Foundation places on "the early development of academic careers dedicated to stimulating the discovery process in which the excitement of research is enhanced by inspired teaching and enthusiastic learning." Two members of the School of Engineering faculty begin the fall semester with the decided advantage of having earned CAREER awards this year. They will both receive substantial financial support, enabling them to continue important research and teaching initiatives.

The grants were awarded to Matthew Begley, an assistant professor of mechanical engineering, and Alexander Shvartsman, an assistant professor of computer science and engineering.

An expert in solid mechanics and material behavior, Begley landed the CAREER Award for work he has done analyzing material properties as they relate to electronic and micro-electro-mechanical devices - devices so small that even the largest of them can sit on a pencil eraser. At the heart of his work is the observation that the actual material properties of very small devices can differ dramatically from the properties of larger devices.

His aim is to develop novel test procedures that will make it possible to analyze a variety of materials, such as polymers and new soldering alloys, used in the production of these miniature devices. Those procedures, which will emerge from laboratory work done by students in existing senior design and accelerated master's degree programs, will employ a nano-indentation facility, a testing device Begley helped develop with a previous NSF equipment grant.

"The point of the CAREER Award is education," says Begley. "This award will enable exceptional learning opportunities for both graduate and undergraduate students. The nano-indenter, for instance, is a tool of the next generation and our students will have many opportunities to employ it in their work, thanks to the National Science Foundation.

"Senior design teams will be creating systems for testing the properties of electronic and micro-electro-mechanical devices," he adds, "work many of them will be involved with as they embark upon their professional careers."

Shvartsman's winning proposal deals with the specification, analysis and development of better functioning distributed systems. His four-year NSF award, beginning this year, includes $200,000 to conduct research and translate his findings into the classroom. Shvartsman's award is complemented this year by an additional $133,000 in NSF funding for research in a related discipline, parallel computing.

Most extant distributed systems, says Shvartsman, include "simple client-server applications where the interaction between system components is confined to the interaction between one client and one server at a time." Examples of such simple systems include the World Wide Web, e-mail systems, bank ATM systems and telephone networks. There are many problems related to emerging demands, however, for which these simple systems are inadequate.

"Distributed banking," he notes, "is a good example, requiring the distribution of accurate information to many different parties, rather than a simple interchange between two parties."

To illustrate the point, he employs a model in which several military leaders must coordinate their action. They may win the battle if they advance on time, but if some of the generals delay, then they will likely lose. If any one of the generals fails to take action in a timely fashion, then the entire battle plan fails.

"If you have a set of processors and they need to agree on a time, and even one is faulty, then the good ones are thwarted," he says. "Our ability to build distributed computing platforms and networks continues to outpace our ability to design sophisticated and reliable distributed applications for these platforms. Despite the existence of many software systems that solve interesting computation problems, and that incorporate advanced algorithms, developing dependable systems for advanced distributed platforms is challenging."

Noting that information management problems will certainly become so complex that distributed systems will be absolutely necessary, Shvartsman said his work will continue to advance the theoretical foundations of distributed systems while exploring the practical implementation of dependable distributed system technology.

Since the CAREER Award strongly emphasizes education, Shvartsman will be engaged in the development of new courses in distributed computing to support both undergraduate and graduate computer science programs, and he will build a research group intended to attract graduate students and postdoctoral researchers.

Jim Smith