Chapter 2. Prehension

Chapter 8 - Constraints on Human Prehension 307
object.’ On the other hand, when a kinesiologist tells a subject to
‘move as quickly and accurately as possible’, these instructions
impose constraints on how to plan and effect the movement, given the
person’s biomechanics and sensorimotor constraints. Even
physicians, when evaluating impairment of hand function, make
distinctions between physical and functional evaluations (Swanson,
Hagert & Scanson, 1983). Having the multiple constraint influences
acting on prehensile systems, roboticists can develop higher level
robot languages for control of more sophisticated dextrous robot
hands, and kinesiologists can further refine their models of human
performance using the tools provided by roboticists.
The constraints shown in Table 8.1 are discussed in the next
section in terms of the three levels of analysis. Then a model is
suggested for capturing human prehensile versatility.
8.1 Physical constraints
The laws of physics, captured in a variety of equations, create
limitations on the planning and control of prehension. Whether the
CNS solves these equations directly or not, they detail the meaning of
movement within inertial reference frames and the effects of making
contact with the environment. In the example where the functional
constraint is to ‘not drop the object’, a posture must be chosen that
effects a stable grasp. The three requirements for a stable grasp are
(Fearing, 1986): 1) the object must be in equilibrium (no net forces
and torques); 2) the direction of the applied forces must be within the
cone of friction, which is twice the angle between the arc-tangent of
the coefficient of static friction and a normal to the surface; and 3) it
should be possible to increase the grasping force’s magnitude to
prevent any displacement due to an arbitrary applied force. Creating a
stable grasp means taking into account active forces and torques as
well as passive ones caused by the frictional components of the skin
contacting the object surface. In addition, counteracting forces must
be separated into their grasping and manipulation components
(Yoshikawa & Nagai, 1990). If the applied force is too powerful,
there is an unnecessary waste of energy; however, what is being
optimized in human systems is a question (Nelson, 1983).
Although useful, sensory information is not a necessary condition
for effecting a stable grasp. In robotics, Mason (1985, 1986) has
shown that there are predictable results when two fingers are about
to grasp an object even though there is some object position uncer-
tainty. For example, in picking up a glass with the right hand, if