These demo programs are part of the Tekkotsu source tree and can be
found in Tekkotsu/Behaviors/Demos/HandEye.
CameraTrackGripper.h
demonstrates use of forward and inverse kinematics solvers. It
relaxes the arm so the user can gently move it around, uses forward
kinematics to claculate the gripper position in body coordinates, and
uses the lookAtPoint function (an inverse kinematics solver) to keep
the camera centered on the gripper.
ArmTrackObject.h.fsm
uses the MapBuilder to find elliptical objects, and the IK solver to
move the gripper to the location of the largest ellipse. This repeats
every two seconds, so the arm will track an easter egg that the user
pushes around the table.
Towers of Hanoi Project from RPI
Students at Rensselaer Polytechnic Institute are using three hand-eye
systems in their spring 2009 Cognitive Robotics course, taught by
Professor Bram Van Heuveln. Evan Patton and Michel Brudzinski created
a Tekkotsu/ACT-R interface that allowed ACT-R to solve Towers of Hanoi
using a real robotic arm. See the project web
page for more information.
RPI doctoral student Michel Brudzinski demonstrating
the Tower of Hanoi solver he and undergraduate Evan Patton developed
as a final project in their cognitive robotics course.
Close-up view of the hand-eye system moving one of the
Tower of Hanoi pieces. These photos were taken during a robot demonstration
at the Schenectady Museum.