Many technology leaders have likened the explosive growth of robotics to the spectacular expansion of the internet in the 1990s. Robots seem to be everywhere and unavoidable. We’ve enlisted them to heal people, explore outer space, aid emergency workers, support soldiers on the battlefield, teach our children, and keep our homes clean. Sales of robots worldwide have grown by double-digit percentages for years. Demand shows no signs of slowing down as we apply these technological marvels to an ever-increasing number of diverse functions. The annual growth in robot sales to some industry sectors, such as electronics and metals, recently exceeded 30 and 40 percent. Overall, forecasters expect the steep upward trend to continue, and Oregon State began riding the curve just as it started to bend sharply skyward.
The seeds of the Oregon State University robotics program within the School of Mechanical, Industrial, and Manufacturing Engineering were planted by Belinda Batten, the former head of the school and currently the executive associate dean of the College of Engineering, who encouraged a group of students to enter the 2005 DARPA grand challenge. From an automobile chassis used in an earlier competition, the participants fashioned a driverless car. Though the small team competed against programs that substantially outspent them, they qualified for the semifinals. That was just the beginning.
Kagan Tumer came to Oregon State in 2006 after a nine-year career as senior research scientist and group lead for the Intelligent Systems Division at NASA’s Ames Research Center. Although individual researchers at Oregon State were already well known for their research in artificial intelligence, computer vision, machine learning, and other disciplines critical to robotics, the robotics program that exists today had not been officially established.
Tumer led the search to hire the first robotics faculty member, Jonathan Hurst, who joined Oregon State in 2008 after completing a doctorate in robotics at Carnegie Mellon University. He was drawn to the new program in Corvallis, because he saw it as fertile ground for something big. “There was very good leadership in the department, and there was support and freedom to build a research group or a program or anything we wanted to do,” he recalled. “I got the sense that the sky was the limit.”
Hurst’s research focuses on the science of legged locomotion, specifically on the passive dynamics of mechanical systems. His group designed and built ATRIAS, a bipedal robot, which has springs in just the right places and legs of just the right shape to take advantage of passive dynamics — — devices that don’t require a power source, such as springs — which enables high-performance running and walking outdoors. Of course, the robot will only function properly with a good control algorithm, so developing software controllers that cooperate with the passive dynamics components was another important research focus for Hurst.
Over the five years following Hurst’s arrival, the robotics program expanded at a fast and furious pace. The college renovated a space for Hurst’s lab, which he named the Dynamic Robotics Laboratory. By 2013, the research group had grown substantially, a trend that continues today.
For incoming professors, the draw of the robotics program includes the chance to work with other departments within the university and take advantage of a priceless opportunity to cross disciplines and exchange ideas with colleagues. “I was really excited to come here,” said Ross Hatton, an assistant professor of mechanical engineering who joined the faculty in 2012. “There were good opportunities for collaborative, interdisciplinary research among the faculty.” Like Hurst, Hatton arrived in Corvallis following his doctoral studies at Carnegie Mellon. He develops fundamental mathematical tools for studying locomotion. His research on spiders and snakes provides the creative foundation for models in which movement is manipulated from a single point of control and incorporates the natural motion of animals into robotic systems. “Biology has many examples of systems that move over rough and complex terrain, and we’ve brought those ideas into robotics,” he said.
Ravi Balasubramanian, an assistant professor who came to Oregon State in 2011, draws inspiration from the human body for designing robotic systems. His cross-disciplinary work brings together surgeons, biomaterials experts, and statisticians to create surgical implants that restore natural movement for people with hand disabilities. His tendon transfer system involves no motors or sensors and, once implanted, is invisible and enables normal hand function. The application isn’t limited to the hand and could be applied wherever tendons connect to muscle.
Bill Smart and Cindy Grimm, both associate professors of mechanical engineering, joined the program in 2012 after spending 11 years on the faculty of the Department of Computer Science and Engineering at Washington University in St. Louis. Smart focuses on human-robot interaction, machine learning, and robotics software. Grimm specializes in the design of robot–human interfaces, computer graphics, and surface modeling. They were both attracted to Oregon State by the broad interdisciplinary nature of the robotics program, which made it easy for them to collaborate with colleagues throughout the university.
Geoff Hollinger, assistant professor of mechanical engineering, joined the team in 2013. His goals include the development of planning, decision-making, and learning techniques that improve robotic systems in the air, on land, and in the ocean. One of his major research thrusts is the development of autonomous underwater vehicles that operate independently of human direction, which requires working at the intersection of robotics, artificial intelligence, and oceanography. “The autonomous robotic systems I design have the potential to revolutionize the way we gather scientific data, improve the efficiency of agricultural production, and even save lives by assisting search and rescue teams,” Hollinger said.
As the master’s and Ph.D. degree programs in robotics began in the fall of 2014, the robotics group continued to work together to recruit students, acquire a shared research space, and build an international standing. Despite the program’s nascent stage, its reputation — and student enrollment — grew rapidly. Just 17 students enrolled in the new degree programs when they were first offered; today, 31 students inhabit the Ph.D. track while 10 pursue master’s degrees, and both have already conferred degrees. While the robotics group rose to an elite level, the number of student applications for advanced degrees rocketed from just over 100 to more than 400. Only a handful of the candidates — the cream of the crop — are accepted. “The international robotics community knows the people we’ve hired,” Hurst said. “In addition, students are very excited about the robotics program, and companies are interested in hiring them.”
In late 2013, Oregon State reached another prestigious milestone when it assumed stewardship of the Robot Operating System (ROS) software infrastructure. ROS is a flexible, open-source framework for writing robotic software that has become the de facto standard in academia and industry and is mandated by a number of well-funded government programs. The OSU Open Source Lab is now the primary hosting site for ROS, supporting an estimated 100,000 users worldwide. This is just one more step toward cementing Oregon State’s position as a robotics hub.
More evidence of the robotic group’s achievements are visible in the 18,000 square-foot, high-bay robotics lab in Graf Hall. It’s open plan, with no physical barriers separating research spaces, stimulates collaboration and interaction among students and faculty researchers.
The high-profile contributions that have emanated from the robotics program reflect the nature of the burgeoning robotics field, according to Hurst. “This growth is a result of a few things,” he said. “First, robotics itself is a growing field with new funding opportunities, new industries, and top faculty candidates making new discoveries in relatively uncharted territory. Second, OSU is growing, and the College of Engineering is responsible for a large percentage of that growth, which necessitates more faculty hiring. Third, the robotics faculty has been successful in creating strong, well-funded research programs. Finally, the school heads and the dean have supported the growth of the robotics group in an effort to build on our success, stay ahead of technology trends, and make the school as a whole stronger.”
Through a combination of strong leadership, top-notch faculty and students, technological innovation, forward thinking, and good timing, Oregon State has blazed a trail to become one of the nation’s strongest academic players in the field of robotics.