Picture of equipment created by ARRI
The Automation & Robotics Research Institute is automating the manufacturing of tiny silicon MOEMS (micro-opto-electro-mechanical systems) for the Office of Naval Research.
"Our body of knowledge is intensive. The knowledge is like Lego blocks, and we must rapidly reconfigure it to other uses."

From large robots to tiny sensors

Automation and Robotics Research Institute embraces second 'Industrial Revolution' to deliver world-class research to customers and transfer technology to industry

Small stuff matters to Harry Stephanou. The director of The University of Texas at Arlington’s Automation & Robotics Research Institute in Fort Worth says miniaturization is the way of the future. No more giant robots assembling automobile chassis.

“That is yesterday’s technology,” Dr. Stephanou says. “You don’t see very many humans on the welding lines in auto plants, and you haven’t in a long, long time.”

He is adamant that the future lies in small things: micro pumps to deliver chemotherapy directly to the affected cells; tiny, inexpensive sensors embedded by the thousands in airplanes to give advance warning when systems are approaching failure; a set of surgeon’s less-than-infant-sized robotic hands that, in addition to the size advantage, never get tremors or experience other human failings.

“There is a second Industrial Revolution under way,” said Stephanou, who notes that because of the shift toward miniaturization, many products must be assembled by humans with flexible hands and small fingers. “We are developing tools that are far smaller than any human hand.”

And when that happens, Stephanou says, the United States will again become a manufacturing economy, using small, inexpensive tools to make small products. “Our body of knowledge is intensive. The knowledge is like Lego blocks, and we must rapidly reconfigure it to other uses.”

Stephanou believes that research exclusively to create new knowledge is also passť. He advocates the more direct approach of asking industries what they need to do the job and then reconfiguring to meet those needs while simultaneously creating new knowledge to anticipate future needs. And if the researcher hits a dead end, that’s what government grants are for: to develop new approaches to solve existing problems.

The ARRI director, who compiled an impressive record in industry before coming to UT Arlington in 2004, intends to have a consortium of companies working with the institute.

One of the institute’s contracts is with the Navy. ARRI produces mechanical locks for munitions, very small sensors that ensure that the torpedo explodes when it is supposed to, not prematurely and too close to the launching vessel. ARRI is doing other work for the Navy, but security concerns prohibit a discussion.

UT Arlington embraced the need for more university-based manufacturing research about 20 years ago when ARRI became a reality. Originally named the Advanced Robotics Research Institute, ARRI’s objective was to combine the resources of the University, Fort Worth/Dallas and relevant industries to introduce manufacturing technologies and strengthen national defense.

Sometimes those new technologies have dictated rather unusual areas of specialization.

Frank Lewis, an electrical engineering professor and holder of the Moncrief-O’Donnell Chair for Automation and Robotics, is a control expert. He came to ARRI in 1990 in the endowed chair dedicated to scholarship and research in the decision-making aspects of artificial intelligence. He has spent more than a decade developing decision-making controls in automated systems that closely resemble human processes.

At the time Dr. Lewis arrived, the National Science Foundation was interested in three aspects of decision making: neural networks, fuzzy logic and discrete event systems. Neural networks are inspired by the human nervous system—the way all the electrons in the body make muscles move. Fuzzy logic is based on the linguistics abilities of humans, and discrete event systems are based on human beings’ decision-making abilities.

Working with computer science and engineering Professor Diane Cook, Lewis developed grant proposals in all three areas. And the researchers received all three types of grants.

“Every faculty member, in essence, runs a small business,” Lewis explains. “One-third is teaching, even though that is the most important part; one-third is research, and the other one-third is bringing in funding.”

Lewis has been successful at each component of his “business.” Awarded numerous grants, he has written five textbooks and dozens of journal papers, and he has graduated 15 Ph.D. students who are taking ARRI research findings to places like Singapore and Croatia. Two of the doctoral graduates won National Science Foundation Career Awards, the highest NSF honor, and others have received awards in South Korea and Ecuador.

ARRI can attract top students, Lewis said, because it provides an opportunity to do industry-relevant work that is mathematically grounded, thus affording returns in both profit and theoretical advancement.

Until the Russians stunned the world in 1957 with the launch of the first artificial satellite, all decision-making controls in the United States were linear, based on a “simplified model of reality.” That worked for technology like radar systems and identifying bombing sites for aircraft, but Sputnik proved that the Russians had embraced nonlinear control theory—a more sophisticated technology that takes into account the interaction and relationships of the system with its environment.

So the United States hurriedly played catch-up and found, in the process, many applications for the new decision technologies.

One of these applications is in the work being done by former graduate student Dan Popa, now a UT Arlington electrical engineering faculty member, on micro-electromechanical systems. Mobile sensor networks can detect fire, intruders, carbon monoxide or any other threat and sound an alarm that dispatches a mobile robot to check the problem.

“It’s a whole system acting together with one goal,’’ Lewis said.

Lewis is working with Texas A&M-Corpus Christi in developing an underwater monitoring system that will be tested in the small lake on the ARRI grounds. The sensors can detect ships, fish, metal and environmental changes and have applications in urgent national priorities like harbor security.

Aside from its research emphasis, ARRI is also home to the Texas Manufacturing Assistance Center, the Cross Timbers Procurement Center and the Small Business Development Center for Enterprise Excellence, maintaining its early commitment to improving the efficiency of local manufacturers.

That’s no small stuff.

visit: http://arri.uta.edu

— Sue Stevens