Abstract
Successful locomotion over complex terrain such as a rocky trail requires walkers to place each step on a safe foothold with a high degree of precision while also taking the future terrain into account. In a previous study presented at VSS 2012, subjects walked over a path strewn with virtual obstacles while we manipulated the distance ahead of the observer at which obstacles became visible. We found that two step lengths of visual information about the upcoming terrain was sufficient for subjects to walk with the same level of performance as they do when vision is unconstrained. Analysis of subjects’ center of mass motion in the different visibility conditions revealed that two step lengths of visual look-ahead provided subjects with sufficient information to exploit their biomechanical structure as efficiently as they do with full vision. In a new experiment, we recorded the accuracy of stepping behavior as subjects walked over a path of irregularly spaced virtual targets. In various conditions, each target became invisible at some point before the subject landed on it. We found that rendering a target invisible at any point during the step to that target (StepN) had no effect on stepping accuracy, but accuracy degraded dramatically if the target became invisible during the preceding step (StepN-1). This finding provides strong evidence that the visual control of stepping to a target occurs during the ballistic swing phase of the previous step; visual feedback of the target during the step itself is unnecessary. This finding supports and advances the theory developed in the previous study that humans use visual information about the upcoming terrain to configure each step so as to exploit the basic biomechanics of bipedal gait and approximate the level of energetic efficiency achievable when walking over flat, obstacle-free terrain.
Meeting abstract presented at VSS 2013