[Apologies for multiple postings] Robotic Motion Planning, Re-planning, and Parallelization GRAND Seminar <http://cs.gmu.edu/~robotics/pmwiki.php/Main/GrandSeminar> Tuesday, November 11, 12 noon, Room 4201 Michael Otte <http://www.mit.edu/~ottemw/> Postdoctoral Associate MIT *Abstract* Reliable sensors, open-source software, and the steady march of Moore’s law have finally placed robotic technologies within the reach of the average citizen. However, we are still searching for motion planning and coordination algorithms that allow robots to move and interact with the same speed, grace, and harmony as humans. My goal is to make the latter a reality, and this talk will highlight some of my work on re-planning in dynamic environments and parallelization of algorithms. In particular, I will talk about RRT-X and C-FOREST. RRT-X is a new Re-planning algorithm that allows real-time navigation in environments that change unexpectedly. Theoretical results derived in the creation of RRT-x lend insight into the two other popular algorithms RRT* (Karaman and Frizzoli) and RRT# (Arslan and Tsiotras). C-FOREST is a parallelization framework for sampling-based motion planning algorithms that achieves significant super-linear speedup in practice (for example, using 64 CPUs yields a speedup over 300X). It has recently been included in the popular The Open Motion Planning Library (OMPL) maintained by the Kavraki Lab at Rice University. *About the Speaker* Michael Otte is a Researcher at the University of Colorado at Boulder in residence at the Aerospace Systems Directorate at the Air Force Research Laboratory, prior to that he was a Postdoctoral Associate at the MIT Laboratory for Information and Decision Systems. He received his PhD in 2011 from the University of Colorado at Boulder. He currently works on the application of artificial intelligence to robotics, with a focus on path planning and multi-agent systems. -- Jyh-Ming Lien Associate Professor George Mason University MASC Group: http://masc.cs.gmu.edu Homepage: http://cs.gmu.edu/~jmlien +1-703-993-9546