Dieter Fox
Assistant Professor, Computer Science and Engineering, University of Washington
Director, Robotics and State Estimation Lab
Faculty Affiliate Researcher, Intel Research Seattle

My main scientific interest lies in the development of computing systems that interact with the physical world in an intelligent way. Such systems will have a tremendous impact on the way we live. Soon, we will live in intelligent environments that are equipped with a variety of sensors and that help their occupants during their everyday lives. We will drive cars that are aware of our goals and that assist us in getting there safely and on time. Mobile robots will support elderly people at home and collaborate with rescue workers in hazardous situations. My goal is to understand the fundamental problems underlying such systems and to develop efficient and robust algorithms based on these insights.

Intel Research Seattle provides many interesting application areas for my research on physically embedded computing systems. Such applications are extremely helpful, since they allow us to evaluate new algorithms and techniques and, even more important, they inspire new research directions by showing us where the most interesting and relevant problems are.

The Intel lab is very helpful in providing us with the infrastructure needed for our research. For example, my work on tracking people using various types of sensors was inspired and made possible by the hardware available in the Intel lab. Ultimately, this work led to more advanced research using GPS sensor data to estimate a person's outdoor transportation routines. The recent paper on this technique that I co-authored with Henry Kautz and Lin Liao won the best paper award at AAAI-04, the main conference in artificial intelligence. Even though this work was not performed in direct collaboration with the Intel lab, it was clearly inspired by our work at the lab.

My student Julie Letchner and I collaborate with Intel Research Seattle on the Place Lab project. (Julie is also an intern at the lab.) This project gives us an opportunity to do research in probabilistic location estimation and sensor modeling. I am very excited about the potentially huge scope of the project, since it is likely to generate a community of thousands (or millions) of users and contributors. This project also nicely integrates with my work on outdoor location and activity recognition.

Along with my university colleagues and students, I also collaborate with the lab on the Human Activity Recognition project. Our research involves activity recognition using sensors such as RFID [radio frequency identification] tags. Specifically, we develop novel, probabilistic techniques for activity recognition using raw sensor data. The really exciting aspect of this project is the benefits it might generate for the health care sector in general and elderly people in particular.

Intel Research Seattle has hired excellent researchers. In addition to collaborating with us on projects, they help us in advising students, and we all benefit from each other's experience in joint seminars. It's great to have the lab and these top-notch researchers so close to campus.