Biomedical Engineering Reference
In-Depth Information
as well as yaw angle. The control algorithm mentioned in the former section was
implemented and successfully applied for the presentation of virtual staircases.
9.4.3 Control Algorithm of the GaitMaster
The control algorithm is required to keep the position of the walker at the neutral po-
sition of the GaitMaster. In order to keep the position constant, the motion-platforms
have to cancel the motion of the feet. The principle of the cancellation is:
(1) Suppose the right foot is at the forward position and left foot is at the backward
position while walking.
(2) When the walker steps forward with the left foot, the weight of the walker is
placed on the right foot.
(3) The motion-platform of the right foot goes backward in accordance with the dis-
placement of the left foot, so that the central position of the walker is maintained.
(4) The motion-platform of the left foot follows the position of the left foot. When
the walker finishes stepping forward, the motion-platform supports the left foot.
If the walker climbs up or goes down stairs, a similar procedure can be applied.
The vertical motion of the feet is canceled using the same principal. The vertical
displacement of the forward foot is canceled in accordance with the motion of the
backward foot, so that the central position of the walker is maintained at the neutral
height.
The turntable rotates so that the two motion-platforms can trace the rotational
motion of the walker. If the walker changes direction of walking, the turntable rotates
to trace the orientation of the walker. The orientation of the turntable is determined
according to direction of the feet. The turntable rotates so that its orientation is at the
middle of the feet. The walker can physically turn around on the GaitMaster using
this control algorithm of the turntable.
A usability test of the prototype GaitMaster was conducted by examining the be-
havior of novice users. Data were collected at the Emerging Technologies venue at
the SIGGRAPH'2000 (July 23-28, 2000, New Orleans). Participants of our demon-
stration experienced virtual flat terrain as well as staircases.
The major interest of this pilot study related to tracking performance of the motion
platforms. Safety straps were put at the level of the foot pad in case the motion
platform failed in tracing the trajectory of the foot.
Through the course of empirical observations, tracking performance was found
insufficient. The position sensor using strings has 0.3 s time delay. This delay implied
an offset between the foot and pad. Walkers often unwittingly stepped off from the
pad. However, the safety straps worked well enough that their feet didn't fall down
from the pad. The toe strap played the main roll in safety. The heel strap was therefore
removed on the 3rd day of the conference. None of the participants experienced any
dangerous situation through the 6-day conference.
