The IEEE Robotics & Automation Society has announced Allen School professor Dieter Fox as the recipient of a 2020 RAS Pioneer Award in recognition of his “pioneering contributions to probabilistic state estimation, RGB-D perception, machine learning in robotics, and bridging academic and industrial robotics research.” The society will formally honor Fox, director of the University of Washington’s Robotics and State Estimation Laboratory and senior director of robotics research at NVIDIA, during the International Conference on Robotics and Automation (ICRA 2020) next week.
The RAS Pioneer Award honors individuals who have had a significant impact on the fields of robotics and automation by initiating new areas of research, development, or engineering. Fox’s contributions have focused on enabling robots to interact with people and their environment in an intelligent way, with an emphasis on state estimation and perception problems such as 3D mapping, object detection and tracking, manipulation, and human activity recognition.
“We are extremely proud that Dieter has been recognized with this prestigious award. It is truly deserved,” said professor Magdalena Balazinska, director of the Allen School. “And it is wonderful to see that both his groundbreaking academic research and his leadership of industrial research labs are being recognized.”
During his career, Fox has earned multiple best paper and test of time awards from major robotics, artificial intelligence, and computer vision conferences that showcase the broad impact he has had on the field. Among his many honors are back-to-back Classic Paper Awards from the Association for the Advancement of Artificial Intelligence (AAAI), which recognizes papers deemed to have been the most influential within the field of artificial intelligence from a given year.
Fox received the first of these honors in 2016 for work he did as a Ph.D. student on “The Interactive Museum Tour-Guide Robot.” Originally published in 1996, that paper introduced RHINO, an autonomous, interactive robot designed to entertain and assist the public in populated environments. RHINO incorporated a number of innovations related to localization, mapping, collision avoidance, and planning to enable it to navigate under uncertainty in challenging conditions — in this case, providing interactive tours to members of the public at the Deutsches Museum in Bonn, Germany. A key aspect of RHINO that set it apart from most other robotics projects at the time was its robust navigation and user interface, which Fox and his colleagues took great pains to make intuitive and user-friendly to non-experts.
The following year, Fox was once again recognized by AAAI, this time for the 1999 paper “Monte Carlo Localization: Efficient Position Estimation for Mobile Robots” published during his time as a postdoctoral researcher at Carnegie Mellon University. Monte Carlo Localization, or MCL, was a new, sample-based algorithm that introduced the use of randomized samples to represent a robot’s belief about its location in a given environment. Fox and his colleagues were the first to apply sample-based estimation in robotics, which they demonstrated to be more accurate, efficient, and easy to use compared to previous approaches such as Kalman filtering based techniques. Among the real-world settings Fox and his colleagues chose to demonstrate MCL was the Smithsonian Museum of Natural History in Washington, D.C., with the help of a robot named Minerva. The team’s sample-based approach has since become the norm for a wide range of applications in the field. A related paper describing MCL that Fox and his colleagues originally presented at ICRA 1999 earned the 2020 IEEE ICRA Milestone Award, which recognizes the most influential ICRA paper published between 1998 and 2002, at this year’s conference.
Robotic vision is an area in which Fox has repeatedly advanced the state of the art during his time at UW. In 2015, he teamed up with postdoc Richard Newcombe and professor Steve Seitz of the Allen School’s Graphics & Imaging Laboratory (GRAIL) to develop DynamicFusion, the first dense simultaneous localization and mapping (SLAM) system for reconstructing dynamic scenes in real time. The project earned Fox and his colleagues a Best Paper Award at the Conference on Computer Vision and Pattern Recognition (CVPR 2015). Two years later, Fox, Newcombe and Ph.D. student Tanner Schmidt earned a Best Robotic Vision Paper Award at ICRA 2017 for presenting “Self-supervised Visual Descriptor Learning for Dense Correspondence.” Leveraging dense mapping techniques such as the aforementioned DynamicFusion, Fox and his collaborators devised an approach for automating the generation of training data and enabling robots to learn the visual features of a scene in a self-supervised way. The project represented a significant leap forward in robot learning by providing a framework for robots to understand their environment without human intervention.
Demonstrating the cross-cutting nature of his work, Fox has earned recognition beyond core robotics and AI conferences. For example, he earned a 10-Year Impact Award from the Association for Computing Machinery’s International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp 2013) for “Inferring High-level Behavior from Low-level Sensors,” which he, along with Ph.D. students Donald Patterson and Lin Liao and then-UW professor Henry Kautz, originally presented in 2003. In their winning paper, the team described a new predictive model of human behavior — in this case, a traveler moving through an urban environment using multiple modes of transportation — using a probabilistic framework that fuses historic low-level sensor data with common-sense knowledge of real-world constraints. Fox and those same colleagues previously earned the inaugural Prominent Paper Award from AI Journal in 2012 for “Learning and inferring transportation routines.” That paper, which was originally published in 2007, built upon the team’s earlier work at Ubicomp by introducing a hierarchical Markov model capable of learning and inferring a user’s daily movements.
A Fellow of both the IEEE and the AAAI, Fox has published more than 240 technical papers on a range of topics and co-authored the textbook “Probabilistic Robotics.”
“For eight years, Dieter was our only robotics faculty member,” said Allen School professor Ed Lazowska, who recruited Fox to Seattle as chair of what was then the UW Department of Computer Science & Engineering. “Today, UW is a powerhouse in robotics research, but it was Dieter who initially put us on the map.
“He is also a monster on a bicycle — none of us can keep up with him,” Lazowska continued, an avid bicyclist himself. “You could say the same for his research. Dieter’s h-index, a measure of the influence of his research publications, is the highest of any of our faculty. It really is impossible to overstate his impact.”
Fox joined the UW faculty in 2000 after obtaining his Ph.D. from the University of Bonn in Germany and completing a postdoc at Carnegie Mellon University with robotics pioneer Sebastian Thrun. He later combined academic research and teaching when he became director of Intel Labs in Seattle and helped establish and co-led the Intel Science and Technology Center on the UW campus. Fox once again bridged academia and industry in 2017, when he took on a position at NVIDIA to start a robotics research effort. In January 2019, he joined NVIDIA CEO Jensen Huang in celebrating the grand opening of the company’s Seattle research lab near the UW campus.
Fox is one of two researchers selected to receive the RAS Pioneer Award this year. Lydia Kavraki, a professor at Rice University and director of the Ken Kennedy Institute, is being honored for “pioneering contributions to the invention of randomized motion planning algorithms and probabilistic roadmaps.” Kavraki and Fox will be recognized during the virtual RAS Society award ceremony on June 5th.
“Being recognized with this award by my research colleagues and the IEEE society is an incredible honor,” Fox said in a related NVIDIA announcement. “I’m very grateful for the amazing collaborators and students I had the chance to work with during my career. I also appreciate that IEEE sees the importance of connecting academic and industrial research — I believe that bridging these areas allows us to make faster progress on the problems we really care about.”