Drug companies apply automation techniques to pill-feeding mechanisms.

Many people hearing the word “automation” picture robots assembling cars in a factory. But an IEEE quarterly journal has, since its introduction almost three years ago, been making every effort to establish automation as a science in its own right and a field separate from the robotics in manufacturing plants.

The journal, Transactions on Automation Science and Engineering (T-ASE), is out to give automation greater visibility—and credibility. And preliminary readership figures indicate that it is succeeding.

DYNAMIC DUO Automation and robotics have often been confused, notes IEEE Fellow Peter B. Luh, professor of electrical engineering at the University of Connecticut at Storrs. Research in robotics today deals mainly with applying intelligent systems to explore the unknown, be it on the ocean floor or on a far-off planet. Because scientists don’t know what will be encountered, robots must be flexible when it comes to the tasks they can perform. Certainly, the accomplishments of the Mars rovers Spirit and Opportunity publicized in the general press indicate that robotics gets more glamorous chores than automation, as well as more media attention and, perhaps because of this exposure, the lion’s share of research funds.

Automation research, in contrast, tackles predetermined tasks, devising systems that repeat a process over and over. Principal concerns are not with the unknown but with speed, precision, efficiency, reliability, quality, and cost-effectiveness. Yes, automation may be incorporated into robotic systems, “but when it works well, you don’t see it,” Luh says. “That very invisibility hampers research, because automation ‘gets no respect,’ to quote comedian Rodney Dangerfield, and it’s hard to attract the best minds here.” This is so “even though many fundamental scientific and practical questions about automation are still unanswered,” he adds.

To lift automation out of the shadow of robotics and address its unique issues—especially the need to set fundamental automation theory on a sounder footing—the IEEE’s Robotics and Automation Society launched its new journal in June 2004. Actually, the RAS split its journal, IEEE Transactions on Robotics and Automation (T-RA), into two publications: T-ASE and IEEE Transactions on Robotics (T-RO).

ATTRACTING ATTENTION “Our goal was to establish T-ASE as the most-cited journal devoted to automation by publishing original, significant, and visionary papers describing new theory and applications,” says Luh, the journal’s editor in chief.

Preliminary numbers of IEEE Xplore’s digital library downloads per paper already indicate that research reported in T-ASE is as sought after as that of other IEEE journals in Xplore that were also launched around 2004. (When this article went to press in February, relevant citation figures were not available. The index used by academic journals to ascertain their importance in a field, tallying what papers from 2004 and 2005 were cited in papers published in 2006—Journal Citation Reports, published annually by Thomson Scientific—had not yet been published for 2006.)

T-ASE is also trying to attract attention from the news media. In November, IEEE Fellow Kenneth Y. Goldberg, who chairs the journal’s advisory board, did a radio and podcast interview called “Automating the World” on the CBS News Radio Network. Goldberg, a professor of industrial engineering at the University of California at Berkeley, discussed the challenges the field faces and some of the advances that the journal has covered.

“When an IEEE journal is founded, it tends to legitimize a field and crystallize a new research area,” Goldberg says. That’s exactly what happened with robotics two decades earlier with the founding of T-RA. He says he hopes the same will happen with T-ASE, especially for encouraging research on the fundamental theories and principles behind automation.


MAJOR CHALLENGE One example of a major unsolved fundamental challenge in automation is parts feeding. If an automated assembly machine is fed a box of randomly oriented parts—brackets, for example—how can it consistently insert each piece into an assembly coming down a production line? Parts feeding is also an issue in the pharmaceutical industry, where one concern is how to funnel millions of pills into hundreds of thousands of tiny bottles without damaging the tablets.

Most factories now solve the parts-feeding problem with custom-built machines. “There’s a whole cottage industry of gurus who devise custom solutions for specific parts,” Goldberg says.

More useful, however, would be a general algorithm that takes a digital model of the part and, without human intervention, develops the specifications for an interface that would orient and feed the parts to the assembly machine. But that requires uncovering mathematical principles for analyzing the geometry, friction, and kinematics of parts of any shape and then figuring out how to get them all to fall in just one orientation. That is the type of fundamental challenge the T-ASE editors are encouraging journal authors to address.

NOT THE FACTORY ALONE Another goal of Luh, Goldberg, and other leaders of the Robotics and Automation Society is to do away with the perception that automation is used only in factories. Automation is also fundamental to monitoring systems (for home and office security and environmental safety), speech recognition (think of directory assistance for telephone numbers), and the task of running hundreds of standard but complex chemical tests to discover new pharmaceutical products. In short, “automation is everywhere,” Goldberg points out.

To home in on these diverse applications, articles in T-ASE have explored new fields. Automating the cultivation of biological cells and the analysis of human DNA was covered in a special issue in April 2006, “Automation for the Life Sciences.” The July 2006 special issue, “Nanoscale Automation and Assembly,” addressed pressing questions about manipulating nanoscale materials by various means, including developing nano-size servo-motors and sensors.

Special issues slated for this year and next include one on systems for automating the home and another dedicated to drug delivery—that is, automating the processes by which medication is released into the body.

FOR MORE INFORMATION about Transactions on Automation Science and Engineering (T-ASE) and the upcoming Conference on Automation Science and Engineering from 22 to 25 September, visit http://www.ieee.org/t-ase.

Goldberg developed an algorithm for rotating any two-dimensional shape into a consistent orientation. Give it a try using an interactive Java applet he has put at http://goldberg.berkeley.edu/part-feeder.