Steve Collins, Steven H. Collins, S.H. Collins, Dynamics, Mechanics, Locomotion, passive-dynamics, Simulation, Controls, human, anthropomorphic, neuromuscular control, musculoskeletal system, simple models, nonlinear controls, foot placement control, Balance, arms, Arm Swinging, Tad McGeer, University of Michigan, Art Kuo, Cornell University, Andy Ruina, nonholonomic booboobechu.
|
Computer simulations are often used to evaluate proposed neuromuscular coordination strategies. It is often appealing to try to include "all" of the complexities of the human body to make a model as "real" as possible. However, having complicated models causes two basic problems. First, computational power requirements explode exponentially with complexity, even to the point of requiring supercomputers (e.g. Anderson and Pandy 1991). Second, and more fundamentally, it can be harder to learn things from complicated models. An overly-complicated model can exhibit overly-complicated behavior, which can be even harder to interpret than experimental data. At their best, models should be abstractions that capture only the most important information. |
|
Simple models of human biomechanics have been forwarded by many, notably R. McNeill Alexander (1995) and Thomas McMahon (1980), and have proven to be very useful for gaining insight into how humans coordinate their motions. In the late 1980's, Tad McGeer suggested an elegant passive-dynamic model for studying human locomotion. |
|
Passive-dynamics presupposes that the motion we understand as walking may in fact largely be the natural dynamical consequence of the mass properties of the human body. That is, people may mostly let their legs swing as they would on their own, then add a little control and power, yielding a gait with inherently low energetic and control demands. Further, McGeer found that very simple models of the human body can exhibit walking motions that are not only remarkably similar to human gait, but also inherently stable without any active control (McGeer 1990). McGeer's original work has inspired and guided those of us in the "passive-dynamics community". My past advisor, Andy Ruina, my current advisor, Art Kuo, their students, and I all use simple, passive-dynamics based models to study human walking. Models of this type have served as inspiration for our walking robots and foot prostheses. I have also developed new passive-dynamic models to study arm swinging and lateral balance through controlled foot placement. (For more on passive dynamics, try one of McGeer's articles or the web pages at U of M, Cornell, or U. Alberta) |
A 3-D Passive-Dynamic Walking Model with Hip-Arms
Left one: model without arms. Center two: fixed arms. Right Three: free-swinging arms. Click images to enlarge.
|
This walking simulation is aimed at exploring the role of arm swinging during walking. My interest in arms began with the first walking robot I built, which pirhouetted wildly until we added arms. Elftman (1939) proposed that humans may use their arms to balance the moment of inertia about a vertical (yaw) axis during gait. To further study this effect, we have a 3-D passive-dynamic walking model with accelerated leg swing, push-off powering (similar to the model of Kuo 1999) and a variety of arm types. Our first aim was to verify that models and robots of this type may have a tendency to spin in yaw without arms, and that the addition of fixed hip-arms, as with our robots, would solve this problem. We then wanted to see whether or not free-swinging hip-arms would naturally entrain to move out of phase with the legs, and identify possible roles of active control of arm torques. (Here, hip-arms are simply arms that share an axis with the legs instead of being at the top of a torso. This is for simplicity, and the hip-arms may be thought of as having properties of both arms and the torso.) The results of this simulation may help lend insight into how and why humans coordinate their arm motions during walking, and we have some human-subject experiments lined up to test these new insights. This work is in progress. |
Interestingly, Martijn Wisse, who worked on the same robot that inspired this model, also became interested in the role of the upper body in reducing yaw-inducing torques, and independently published a paper on the matter. |
Active Control of Lateral Balance through Foot Placement
Left: movie of unstable 3-D walking. Right: movie of stabilized walking. Click images to watch movies.
|
This walking simulation is based on a similar simulation by Kuo (1999), which showed that a 3-D model required active control to maintain lateral balance as it walked. To a similar passive-dynamic model, we have added an internal-model based controller that provides lateral stability through lateral foot placement. The controller actuates lateral foot placement based on an internal estimation of the model's state that is obtained by integrating noisy sensory information from simple models of the human otoliths, semicircular canals, eyes, and proprioceptive organs. The external model is also perturbed by random noise, both continuously and discretely at heel strike events. Thus, we are able to compare foot placement variability with our expectations and with human subject data (collected by Catherine Bauby (2000)). Further, we are able to simulate aging and the closing of one's eyes by varying the sensory inputs and sensory noise, which allows us to make predictions for the results of human subject experiments on young and elderly individuals. This work is in progress. |
Proceedings: Collins, S.H., Bauby, C.E., Kuo A.D. (2003) Control of Balance During Walking in Young and Elderly Adults. In Proc. 27th Ann. Mtg. American Society Biomechanics, Toledo, OH. |
This work has been funded by the University of Michigan and a National Aeronautics and Space Administration (NASA) Graduate Student Researhers Program (GSRP) grant, through the Johnson Space Center.
Last updated: 2/19/2005
References
Alexander, R. M. (1995) Simple Models of Human Motion, Applied Mechanics Reviews, 48, 461-469. | ||
Alexander, R. M. (2003) Principles of Animal Locomotion. New Jersey: Princeton University Press. | ||
Anderson, F.C., Pandy, M.G. (2001) Dynamic optimization of human walking. J. Biomechanical Engineering 123: 381–390. | ||
Bauby, C. E., and Kuo, A. D. (2000) Active control of lateral balance in human walking. Journal of Biomechanics, 33: 1433-1440. | ||
Coleman, M., Chatterjee, A., Ruina, A. (1997) Motions of A Rimless Spoked Wheel: A Simple 3D System With Impacts. Dynamics and Stability of Systems, 12(3): 139-160. | ||
Coleman, M. (1997) A stability study of a three-dimensional passive-dynamic model of human gait. Ph.D. Thesis, Cornell University, Ithaca, NY USA. | ||
Coleman, M., Garcia, M., Mombauer, K., Ruina, A. (2001) Prediction of Stable Walking for a Toy That Cannot Stand Still Physical Review E, 64: 2 | ||
Donelan, J. M., Shipman, D. W., Kram, R., and Kuo, A. D. (2004) Mechanical and metabolic requirements for active lateral stabilization in human walking. Journal of Biomechanics, 37: 827-835. | ||
Elftman, H. 1939. The function of arms in walking. Human Biology 11:529–535. | ||
Garcia, M. (1998) Stability, scaling, and chaos in passive-dynamic gait models. Ph.D. Thesis, Cornell University, Ithaca, NY USA. | ||
Garcia, M., Chatterjee, A, Ruina, A. (2000) Efficiency, speed, and scaling of two-dimensional passive-dynamic walking. Dynamics and Stability of Sytems, 15:2 75-99 | ||
Garcia, M., Ruina, A., Chatterjee, A., and Coleman, M. (1998) The Simplest Walking Model: Stability, Complexity, and Scaling. J. Biomechanical Engineering, 120(2): 281-288. | ||
Kuo, A. D. (1999) Stabilization of lateral motion in passive dynamic walking, International Journal of Robotics Research, 18 (9): 917-930. | ||
Kuo, A.D. (2001) A simple model of bipedal walking predicts the preferred speed-step length relationship, Journal of Biomechanical Engineering, 123: 264-269. | ||
Kuo, A. D. (2002) Energetics of actively powered locomotion using the simplest walking model, Journal of Biomechanical Engineering, 124: 113-120. | ||
McGeer, T. (1989) Powered flight, child's play, silly wheels, and walking machines. In: Proc. IEEE Robotics and Automation Conference, Piscataway, NJ : pp. 1592-1597. | ||
McGeer, T. (1990) Passive dynamic walking. International Journal of Robotics Research. 9(2): 68-82. | ||
McGeer, T. (1990) Passive dynamic running. Proceedings of the Royal Society of London B. 1240: 107-134. | ||
McGeer, T. (1990) Passive walking with knees. In: Proc. IEEE Robotics and Automation Conference, Cincinnati, OH : pp. 1640-1645. | ||
McGeer, T. (1991) Passive dynamic biped catalogue. In: Proc. 2nd Int. Symp. of Experimental Robotics, Toulouse, France: pp. 465-490. | ||
McGeer, T. (1992) Principles of walking and running. In Advances in Comparative and Environmental Physiology, Vol. 11, Mechanics of Animal Locomotion (Alexander, R. M., ed.). Berlin: Springer-Verlag | ||
McGeer, T. (1993) Dynamics and control of bipedal locomotion. Journal of Theoretical Biology. 163: 277-314. | ||
McMahon, T. A., Bonner, J.T. (1983) On Size and Life. New York: Scientific American Books. | ||
McMahon, T. A. (1984) Muscles, Reflexes, and Locomotion. New Jersey: Princeton University Press. | ||
Mochon, S., McMahon, T. A. (1980) Ballistic Walking: An Improved Model. Mathematical Biosciences, 52: 241-260. | ||
Wisse, M., Schwab, A. L. (2001) A 3D Passive Dynamic Biped with Yaw and Roll Compensation. Robotica, 19(3) 275-284. |