Nov 17 (Thursday) John Bertram (University of Calgary)

A minor epiphany in gait control: understanding constrained optimization

Abstract:

Humans will spontaneously select different step length ­ frequency
relationships in walking depending on externally applied constraints (as
in set speed, frequency or step length). Such behavior can be accounted
for as a classic constrained optimization solution where cost of transport
is the objective function. Unusual gait parameter combinations are
required to minimize transport cost, yet these are perceived as quite
‘natural’ under the circumstances. Having made this observation, it is of
interest to consider the process through which the minimization function
is determined and implemented by the individual. Since the physiological
response to activity has a substantial delay, it is unlikely that central
monitoring of byproducts of metabolic cost would be adequate to guide the
system. Alternatively, activity in the peripheral components that
ultimately contribute to the overall cost are complex and dynamic.
Although logically possible, the acquisition of sensory input from the
activity of all components and the integration of a functional strategy to
minimize cost seems an onerous task on a step-by-step basis. However,
humans (and presumably other animals) have a highly developed sense of
task effort that derives from the motive excitation of the musculature
(motor outflow commands). This provides an opportunity to immediately but
indirectly evaluate the relative ‘cost’ of activity. It is expected that
the apparent cost indicated by such a system could be closely related to
real cost for some situations, i.e. walking and running, but may be
confounded in circumstances in which overall performance is not as
directly related to apparent immediate effort, i.e. cycling. In this
seminar constrained optimization in a variety of activities will be
compared and discussed in light of potential control mechanisms.