A postdoc position is available in the Intelligent Motion Lab (IML) at Duke University in the area of motion planning using experience. An emerging direction of research is to generate massive amounts of virtual experience for robots in simulation, enabling "big data" approaches to learning optimal robot motions and feedback control policies. Our lab has been exploring this approach, and have set a goal of performing optimal motion planning for complex, high dimensional systems in real-time.
The postdoctoral researcher on this project will conduct a vigorous research program in collaboration with Prof. Kris Hauser and his graduate students and external collaborators. The candidate will make theoretical contributions at the intersection of motion planning and machine learning to achieve performance improvements beyond the existing state-of-the-art. Successful candidates will have demonstrated a track record of high-quality academic publications and scholarly engagement with the robotics, AI planning, and/or optimization community. Ideal candidates will also have strong communication skills, leadership skills, and an interest in robotics education. Contributions to open-source software projects and research experience in application areas (e.g., intelligent vehicles, human-robot interaction, legged robots) are also a plus.
Many opportunities are available for the candidate to apply his/her research to real-world applications on robot hardware platforms, such as the Amazon Picking Challenge, mobile manipulation with the TeleRobotic Intelligent Nursing Assistant (TRINA), and rock-climbing with the JPL RoboSimian quadruped robot.
This position is open from June 2015 with funding guaranteed for one year, with the possibility of renewing for one or more subsequent years. Salary is competitive. To apply, please email a CV to Kris Hauser at: email@example.com.
More information about the lab can be found at http://motion.pratt.duke.edu.