Catching a free-falling object by hand is a taks often used to study the capabilities
of the central nervous system.
The main problem regarding the completion of this task is to move the hand to
the right place at right time.
To solve this problem with robots, tangential catching (i.e. matching of velocities
between manipulator and object) is chosen to reduce the degrees of freedom of
this task and to allow much greater robustness to the exact timing of the hand
closure (Hong and Slotine, 1995).
The reliability advantage of side catching becomes even more important when
targets are more elongated in the motion direction, since it makes it easier
to stabilize the target in the hand and makes exact timing of the closing even
less important.
In line with those observations, we analyzed hand kinematics during a free-falling
catching task in humans.
Six healthy subjects were asked to catch an iron bar dropped by the experimenter (from 2 different heights) or by himself (1 height), with each hand. Bar and hand kinematics were captured by means of an optical 3D-system.
Based on several kinematics parameters, we found the following results :
Hand kinematics analysis revealed that subjects used different anticipatory strategies not related to the relative bar/hand position (prospective model) but only to stick position (predictive model). Hand movement components were controlled independently. Morevoer, humans used a strategy similar to that implemented by robots that simplify the control of the movement by reducing 3D catching task to a catching in 1D.
Hong W. & Slotine J.-J.E. (1995). Experiments in hand-eye coordination using active vision. Experimental Robotic IV, Sprinter-Verlag, Proceeding of the 4th International Symposium on Experimental Robotics, ISER’95, Stanford, Califonia, June 30-July-2.