A future in which robot-powered automated manufacturing lines endlessly pump out products across industries sounds like a boon for efficiency and productivity.
It also feels more than a little dystopian, triggering fears of human workers being replaced by machines at an unprecedented scale.
But from Robert Radwin’s perspective, that vision is neither realistic nor prudent.
Robert Radwin
“Researchers are beginning to realize that a future workplace dominated by robots and completely devoid of human workers is not only impractical but undesirable, for a number of reasons,” says Radwin, a professor of industrial and systems engineering at the University of Wisconsin-Madison who’s studying how to best integrate robots into manufacturing industries. “We envision a workplace where workers won’t be replaced by robots, but rather where robots will assist workers in their jobs. That’s our goal.”
Radwin is working on several projects to devise strategies for introducing collaborative robots (or “cobots”) into manufacturing jobs in ways that enhance work—boosting overall productivity while reducing physical and mental workloads for human workers. To do so, he’s collaborating with companies like Wisconsin-based Mercury Marine, General Motors and Boeing, and teaming up with UW-Madison researchers in industrial and systems engineering, computer science, mechanical engineering and labor economics to consider the technical challenges, ergonomics considerations and employment policy issues.
As opposed to colossal industrial robots that function explicitly separate from human workers because of safety considerations, cobots are designed to operate in concert with humans. They move slower, deliver less brute force and include more sensors to avoid causing injury. Yet they can endlessly perform repetitive motions, relieving physical stress for human workers.
In order to maximize cobots’ potential, though, companies need to strategically integrate them into manufacturing workflows in ways that truly optimize productivity and reduce both physical and mental strain, rather than unintentionally adding complexity. That’s where Radwin and his UW-Madison colleagues come in.
Through a series of National Science Foundation grants, Radwin is leading an effort to apply a data-driven methodology for evaluating jobs for human-robot collaboration—outlined in a March 2022 paper in the journal Human Factors—to real roles at Mercury Marine’s foundry and die-casting plant in Fond du Lac, Wisconsin, and a General Motors assembly line. The research, part of NSF’s Future of Work at the Human-Technology Frontier program, will take a multidisciplinary approach with the help of Professors Bilge Mutlu (computer sciences), Shiyu Zhou (industrial and systems engineering) and Timothy Smeeding (public affairs), and Assistant Professor Lindsay Jacobs (public affairs).
Radwin is also part of a NASA-funded project led by Mutlu to apply a similar approach to aviation manufacturing with industry partner Boeing. Michael Zinn, an associate professor of mechanical engineering, is a collaborator, as is Michael Gleicher, a professor of computer sciences.
“Today employers are challenged to find employees,” says Radwin, who has a long track record of using his ergonomics research to assist industry partners. “Right now, there’s a shortage of workers. It’s natural to think of trying to substitute workers with robots. But we think that, in fact, we can gain productivity and some of the benefits that employers are seeking by creating a better match between existing employees and automation.”
Radwin points to fears in the early 1970s that the introduction of ATMs would eliminate the need for bank tellers. Instead, the number of tellers has grown along with the expansion of bank branches. It turns out, in addition to making branch offices cheaper to operate with fewer tellers, ATMs handled tedious tasks like dispensing cash, freeing up tellers to assist customers with more complex requests—the sort of work humans are still better suited to handle.
“We get gratification from our work,” says Radwin, “and so that’s something that we think is an important aspect of creating jobs of the future that eliminate the bad stuff and enhance the good stuff, and at the same time improve productivity and make jobs better for workers, both physically and mentally.”
Robert Radwin is the Duane H. and Dorothy M. Bluemke Professor of Industrial and Systems Engineering. Bilge Mutlu is the Sheldon B. and Marianne S. Lubar Professor of Computer Science. Shiyu Zhou is a Vilas Distinguished Achievement Professor of Industrial and Systems Engineering. Timothy Smeeding is the Lee Rainwater Distinguished Professor of Public Affairs and Economics. Lindsay Jacobs is an Assistant Professor of Public Affairs.
Top photo caption: Anna Konstant, a PhD student in Robert Radwin’s lab, examines a collaborative robot that could be used in aviation manufacturing. Photo: Joel Hallberg.