Parker C. Lusk
robotics and mobile autonomy • estimation, perception, and control
I am a PhD candidate at MIT under the supervision of Prof. Jonathan How working in robotics and autonomous systems. My research focuses on solving challenging data association problems that frequently arise in fundamental estimation problems (e.g., localization, SLAM, calibration, multiple object tracking, multirobot map merging). I have developed algorithms that enable robust estimation in the presence of many outliers, even without an initial guess. These algorithms leverage graph theory, optimization, and manifold geometry.
With extensive experience making real robots do real things, I have a knack for turning theory into practice. I have nearly a decade of experience in C++/Python for robotics and have been programming for nearly two decades. I am passionate about designing and implementing advanced algorithms and seeing the process through to deployment and evaluation in hardware. My technical expertise spans computer vision, geometry, control, linear algebra, optimization, and optimal estimation. I love sensors, making robots move, probabilistic robotics, navigation, dealing with uncertainty, maximum a posteriori formulations, and saying “on the manifold”.
I received the M.S. in Electrical and Computer Engineering (2018) and the B.S. in Electrical Engineering (2016) from Brigham Young University, Provo, UT, where I was advised by Prof. Randy Beard.
selected projects
selected publications
- CLIPPER: Robust Data Association without an Initial GuessIEEE Robotics and Automation Letters, 2024
- MOTLEE: Distributed Mobile Multi-Object Tracking with Localization Error EliminationIn IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2023
- Global Relocalization in Unstructured Environments using Semantic Object Maps Built from Various ViewpointsIn IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2023IROS Best Paper Award on Safety, Security, and Rescue Robotics, Finalist
- MIXER: Multiattribute, Multiway Fusion of Uncertain Pairwise AffinitiesIEEE Robotics and Automation Letters, 2023
- GraffMatch: Global Matching of 3D Lines and Planes for Wide Baseline LiDAR RegistrationIEEE Robotics and Automation Letters, 2022