Matt Mason
Matt Mason grew up in the suburbs of Oklahoma City, where he spent his time on piano, baseball, math and debate. Although he liked to build things, most of his projects (most notably the Lake Hefner Submarine Initiative) fell short. However, important scientific dividends from such projects, such as the discovery that used tires do not make good flotation devices, set the pattern for his future career.
Mason discovered computers at a Stanford University summer high school program in 1969, and majored in computer science at MIT as an undergraduate. There he spent endless hours playing spacewar on the PDP1, before apprenticing himself to Gerry Sussman at the Artificial Intelligence Lab, leading to his first research project and bachelor's thesis, "A Qualitative Simulation of Swine Production", one of the earliest and least successful attempts ever to apply AI to agriculture.
Mason's interests turned to robotics during his first year in graduate school at MIT, when he saw how much fun his office mates were having. He first worked on how a robot arm can move gracefully while touching a hard surface. This project, Mason's M.S. thesis, is to this day his most-cited work and formed the basis for what is sometimes called "hybrid force/position control."
Struggling in vain to surpass his M.S. work, Mason then began to study pushing, and also helped to develop a theory for automated planning now generally referred to as "LMT" for its three authors Lozano-Perez, Mason and Taylor. Mason wrote his Ph.D. thesis on pushing, and then moved to Carnegie Mellon University in 1982.
Over the years Mason and his students, collaborators and colleagues developed a distinctive research style: (1) pick some specific mode of manipulation, such as compliant motion, pushing, squeezing, or striking; (2) use Newtonian mechanics to determine the likely outcomes of the robot's possible actions; (3) use AI, control engineering, and ad hoc methods to plan and control the robot's actions given the desired outcomes. Along the way they developed a variety of interesting machines and demonstrations: an origami-folding robot, a small wheeled robot that scoots paper around on a desktop, the Palm Pilot Robot Kit, a robot that throws blocks with dazzling dexterity, and even a robot that mimics an inverted turtle that moves on its back by wiggling its legs in the air.
Mason was the chairperson of the robotics Ph.D. program at Carnegie Mellon for many years, and has been the director of the Robotics Institute since 2004. He co-authored "Robot Hands and the Mechanics of Manipulation" (MIT Press 1985) with his colleague Ken Salisbury, and co-edited "Robot Motion: Planning and Control" (MIT Press 1982) with several of his colleagues at MIT. Mason's most recent book "Mechanics of Robotic Manipulation" (MIT Press 2001) is a textbook based on a graduate course he has taught at CMU for many years. Mason is a Fellow of the AAAI, and a Fellow of the IEEE. He is a winner of the System Development Foundation Prize (shared with Ken Salisbury) and in 2009 won the IEEE Robotics and Automation Society's Pioneer Award for "pioneering contributions to the fundamental understanding of the mechanics of robotic manipulation and to graduate education in robotics."
http://www.ri.cmu.edu/people/mason_matthew.html
| tour request | contact irobot | privacy policy | terms & conditions |  |