The consistency of maximum running speed measurements in humans using a feedback-controlled treadmill, and a comparison with maximum attainable speed during overground locomotion.

Consistent measurement of maximum running speed overground is problematic due to the difficulty in precise, continual measurement of speed, and the substantial workload in accelerating the body promoting the onset of fatigue. Treadmills remove the requirement for acceleration which enables more repeats. They also allow experiments to be carried out in controlled environments and where space is limited, but they usually depend on manual and subjective speed control. Here we used a draw-wire position sensor and a proportional-derivative (PD) controller to automatically adjust treadmill belt speed of a large equine treadmill. The feedback loop took the real-time position and velocity of the runner relative to the front of the treadmill as input. This control system allowed runners to accelerate from walking speed to a peak running speed within a few strides and then decelerate as quickly as they wished. We used the system to evaluate the variation in maximum speed determination that results from one trial to 10 trials, in eleven individuals. Three trials gave a maximum speed 97.8% of that achieved after ten. The approach used is appropriate for any treadmill where the running zone length is greater than three metres and the speed controller can be externally controlled. Subjects ran 11.5% faster on the treadmill than overground, part of which can be explained by the removal of aerodynamic drag and the fatigue of overground running. Additional factors may, however, contribute to athletes running faster on a treadmill, for instance some aspect of stability or control.