@import url(http://kodlab.seas.upenn.edu/pub/skins/sinorca/basic.css); @import url(http://kodlab.seas.upenn.edu/pub/skins/sinorca/layout.css); @import url(http://kodlab.seas.upenn.edu/pub/skins/sinorca/sinorca.css);
I was a member of Kodlab from 2008 to 2011, supported by an Intelligence Community Postdoctoral Research Fellowship. Previous to Penn, I received the Ph.D. and M.S. degrees in Robotics from CMU’s Robotics Institute (2008, 2005) and a B.S.E. from the University of Michigan (2002).
At its broadest level, my research focuses upon techniques for coordinating the motions of mobile robots, in particular controlling the legs of multi-legged robots as they locomote over extreme terrain. As such, my work has traditionally involved varied fields such as topological issues of control, effective robot programming, development of high performance robots and robot mechanisms, plus much more.
Since working in a robotics lab as an undergraduate student, on through graduate school, and now to Penn, I have had the pleasure of working with some of the highest performance legged machines ever devised, robots capable of running and climbing through extremely difficult terrains.
During graduate school, I was the chief behavioral designer for the RiSE platform, developing gaits and behaviors that allowed the robot to climb a variety of useful outdoor vertical surfaces (including work on RiSEV1, RiSEV2, and RiSEV3). Along the way, I have also had the pleasure of working on various RHex-style robots (Research RHex, EduBot, and our newest robot in Kodlab, X-RHex), as well as the Little Dog platform.
Robots like these have the potential to travel to terrains and locales previously inaccessible by any robotic system, and my work on behavioral control seeks to make locomotion over these extreme terrains possible.
Topology is the study of underlying “spaces” for a given system, describing basic properties of the system that are preserved despite modifications to that system. In our studies on the topology of gaits, we have identified topological complexes that exist for all forms of gaited locomotion systems, and have devised control methods applicable to all legged systems.
Gaits are defined as repetitive leg motions that produce overall locomotion by the body of a robot or animal. Quadrupeds can crawl, trot, pace, bound, or gallop, generally the fastest quadruped gaits. Hexapedal creatures, such as insects studied by our colleagues, often use crawl gaits and the alternating tripod gait, which is the primary gait used by our RHex robots. In my work, I have considered the full space of all gaits, defined as cycles on an N-dimensional torus, and have identified two very important structures on that space, the Gait Complex and the Stance Complex.
Legged creatures do not have to plan every footfall they make, yet most legged machines still do exactly that. By taking advantage of unique mechanics for robotic mechanisms, it is possible to achieve dramatically good locomotion while solely using open-loop gaits. My Ph.D. thesis attempted to find methods by which we could apply small amounts of feedback control to feedforward gaits in order to improve locomotion by these systems. As such, I have developed several unique ideas on the control of legged locomotion, and have applied these ideas to quadrupedal and hexapedal platforms.
Relevant publications: Thesis
Read about Dynamism, a framework developed here in Kod*lab for use in writing behaviors for real-time robots.
Main page: Publications
On the Comparative Analysis of Locomotory Systems with Vertical Travel
G. C. Haynes and D. E. Koditschek
International Symposium on Experimental Robotics, December 2010.
X-RHex: A Highly Mobile Hexapedal Robot for Sensorimotor Tasks
Kevin C. Galloway, G. C. Haynes, B. Deniz Ilhan, Aaron M. Johnson, Ryan Knopf, Goran Lynch, Benjamin Plotnick, Mackenzie White, D. E. Koditschek
University of Pennsylvania Technical Report, 2010.
Disturbance Detection, Identification, and Recovery by Gait Transition in Legged Robots
Aaron M. Johnson, G. Clark Haynes, D. E. Koditschek
IEEE/RSJ International Conference on Intelligent Robots and Systems, October 2010.
Gait Transitions for Quasi-Static Hexapedal Locomotion on Level Ground
G. C. Haynes, F. R. Cohen, D. E. Koditschek
International Symposium of Robotics Research. August 2009
Rapid Pole Climbing with a Quadrupedal Robot
G. C. Haynes, Alex Khripin, Goran Lynch, Jon Amory, Aaron Saunders, Alfred A. Rizzi, and D. E. Koditschek
IEEE International Conference on Robotics and Automation. May 2009
Integrating a Hierarchy of Simulation Tools for Legged Robot Locomotion
Andrew Slatton , Daniel Cohen, Yang Ding, Paul B. Umbanhowar, Daniel I. Goldman, G. Clark Haynes, Haldun Komsuoglu, and D. E. Koditschek
Workshop on Robot Simulators, IEEE/RSJ International Conference on Intelligent Robots and Systems. September 2008
Gait Regulation Control Techniques for Robust Legged Locomotion
G. C. Haynes
Doctoral Dissertation, Tech. Report CMU-RI-TR-08–19, Carnegie Mellon University Robotics Institute. May 2008
Biologically Inspired Climbing with a Hexapedal Robot
M. A. Spenko, G. C. Haynes, J. A. Saunders, M. Cutkosky, A. Rizzi, R. J. Full, D. E. Koditschek
Journal of Field Robotics. Volume 25, Issue 4–5, 2008
Gait Regulation and Feedback on a Robotic Climbing Hexapod
G. C. Haynes and A. A. Rizzi
Proceedings of Robotics: Science and Systems. August 2006
Copyright Kodlab, 2014