Research Papers

Biomechanics of Step Initiation After Balance Recovery With Implications for Humanoid Robot Locomotion

[+] Author and Article Information
Christine Miller Buffinton

Department of Mechanical Engineering,
Bucknell University,
One Dent Drive,
Lewisburg, PA 17837
e-mail: christine.buffinton@bucknell.edu

Elise M. Buffinton

School of Civil and Environmental Engineering,
Cornell University,
220 Hollister Hall,
Ithaca, NY 14853
e-mail: emb368@cornell.edu

Kathleen A. Bieryla

Department of Biomedical Engineering,
Bucknell University,
One Dent Drive,
Lewisburg, PA 17837
e-mail: k.bieryla@bucknell.edu

Jerry E. Pratt

Florida Institute for Human
and Machine Cognition,
40 South Alcaniz Street,
Pensacola, FL 32502
e-mail: jpratt@ihmc.us

1Corresponding author.

Manuscript received March 9, 2015; final manuscript received December 30, 2015; published online January 29, 2016. Assoc. Editor: Kenneth Fischer.

J Biomech Eng 138(3), 031001 (Jan 29, 2016) (9 pages) Paper No: BIO-15-1105; doi: 10.1115/1.4032468 History: Received March 09, 2015; Revised December 30, 2015

Balance-recovery stepping is often necessary for both a human and humanoid robot to avoid a fall by taking a single step or multiple steps after an external perturbation. The determination of where to step to come to a complete stop has been studied, but little is known about the strategy for initiation of forward motion from the static position following such a step. The goal of this study was to examine the human strategy for stepping by moving the back foot forward from a static, double-support position, comparing parameters from normal step length (SL) to those from increasing SLs to the point of step failure, to provide inspiration for a humanoid control strategy. Healthy young adults instrumented with joint reflective markers executed a prescribed-length step from rest while marker positions and ground reaction forces (GRFs) were measured. The participants were scaled to the Gait2354 model in opensim software to calculate body kinematic and joint kinetic parameters, with further post-processing in matlab. With increasing SL, participants reduced both static and push-off back-foot GRF. Body center of mass (CoM) lowered and moved forward, with additional lowering at the longer steps, and followed a path centered within the initial base of support (BoS). Step execution was successful if participants gained enough forward momentum at toe-off to move the instantaneous capture point (ICP) to within the BoS defined by the final position of both feet on the front force plate. All lower extremity joint torques increased with SL except ankle joint. Front knee work increased dramatically with SL, accompanied by decrease in back-ankle work. As SL increased, the human strategy changed, with participants shifting their CoM forward and downward before toe-off, thus gaining forward momentum, while using less propulsive work from the back ankle and engaging the front knee to straighten the body. The results have significance for human motion, suggesting the upper limit of the SL that can be completed with back-ankle push-off before additional knee flexion and torque is needed. For biped control, the results support stability based on capture-point dynamics and suggest strategy for center-of-mass trajectory and distribution of ground force reactions that can be compared with robot controllers for initiation of gait after recovery steps.

Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

(a) An instrumented participant in static double-stance position prior to stepping. (b) opensim Gait2354 model of a participant at SL1, normal SL. (c) opensim Gait2354 model at SL3, normal SL plus 20% of height. Green arrows are directed from the CoP and show the direction and magnitude of the GRF vector.

Grahic Jump Location
Fig. 2

Components of the GRF for back and front feet during characteristic stepping trials: vertical, A–P, and M–L. Force is normalized to the subject's body weight. (a) SL1, normal SL; (b) SL5, normal SL plus 40% of height. As SL increased, more weight shifted to the front foot in the static position and back vertical push-off force declined.

Grahic Jump Location
Fig. 3

The position of the body CoM in the sagittal plane for three representative participants during five different SLs. Each line is a trace of the CoM with forward (A–P) direction as the abscissa and vertical direction as the ordinate, with time as a parameter on the curve. The CoM position begins at the left end of each curve and follows the direction indicated by the arrows. The horizontal positions are normalized to the starting position at SL1, normal SL. (a) Female, (b) and (c) male.

Grahic Jump Location
Fig. 4

Locations of the CoP, CoM, and instantaneous capture point (ICP) on a ground tracing for three characteristic SLs for a single participant: (a) SL1, (b) SL3, (c) SL5. The points indicated are S: starting static position, TO: toe-off position, and F: final position, when both feet are on the front-force plate. Ground projections showing the outline of each foot based on four markers for the beginning and end of each step are also shown, with the BoS sketched for the static starting position. The path of the CoM is traced in dashed line; the path of the ICP is indicated in red.

Grahic Jump Location
Fig. 5

Characteristic joint torques normalized to body mass in six joints for a single participant during motion at two SLs: (a) SL1, normal SL; (b) SL5, normal SL plus 40% of height. A positive value of torque represents hip and knee flexion and ankle dorsiflexion. Comparison of the two SLs shows large increases in torque in the back hip and front and back knees.

Grahic Jump Location
Fig. 6

Variation of joint kinetic parameters with SL, all normalized to body mass. Dashed line is back leg (stepping leg) and solid line is front leg (stance leg). Results are mean of the averaged trials for each subject at the indicated SL, with male and female participants combined. The error bars indicate the 95% CI. Numbers above or below an error bar indicate the SL(s) with values that are significantly different at p < 0.05 from the indicated SL. (a) Peak moment, (b) peak power, and (c) total work.




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In