|
Of several possible forms of human-robot
interaction, we focus on the case which the human and the robot cooperative
execute a manual task (as opposed to haptic display of a virtual
environment). We are studying the use of programmable constraints for safe
robot-assisted manipulation of heavy loads in materials handling, assembly,
and construction. Frictionless constraints guide the load to the goal,
allowing the human to choose force strategies which minimize the possibility
of injury.
To design programmable guides which increase productivity and safety, we are
studying how humans naturally interact with constraints. Our studies have
shown that subjects consistently apply forces against the constraint, even
though these forces have no effect on the task. This implies the existence of
force-direction preferences, depending on the body configuration. Without any
detailed biomechanical modeling or psychophysical questioning, we have been
able to derive this preference directly from data in a planar single-arm
task.
Next, a 2-dof cobot, which presents stiff yet smooth constraint paths to a
human user grasping the handle, will be used to realize programmable
constraints. The wide, low design allows large constraint forces and a full
range of human arm motion. The practical applications of our study will be
offering a safe, passive alternative for robotic rehabilitation and robot-assisted
materials handling.
Videos
- Enforcing
Virtual Paths (10.6 MB)
The UTLA can constrain a user to almost any path. This video illustrates
a user constrained to the path of a figure eight. As the user is moving
along the figure eight, the path is changed to a circle.
- Dynamic
Effects (13.8 MB)
As a passive device, the dynamics of the UTLA are apparent to the user.
This video demonstrates the effects of the dynamics on a user's motion.
The user is constrained to a circular path. When the user is at the top
and bottom of the path, only the second link of the UTLA must be moved.
When the user is at the left and right side of the circle, however, both
links must be moved in order for the end-effector to move. This means
that for a specified power input at the top and bottom, the end-effector
will move faster than for the same power input at the left and right
sides. Thus as the user moves around the circle, the end-effector will
speed up and slow down.
- The
Steered Wheel (21.3 MB)
The UTLA, operating in Virtual Caster mode, steers its wheel to allow
motion in the direction indicated by a user's force vector. A close-up
of the wheel is shown, including a view of the traction drive
transmission, the motor, and the force sensor.
- Reaching
Demo (2.7 MB)
One of the possible applications for the UTLA is as a tool for stroke
rehabilitation. A patient can be asked to perform a repetitive reaching
task through the UTLA's workspace. While reaching, if the UTLA detects
that the user is veering off-course, it can actively correct the user's
direction of motion.
This video illustrates a healthy user reaching along various paths
through the UTLA's workspace. The paths are altered after each reaching
motion is complete.
Publications:
Tanya Tickel, David Hannon, Kevin M. Lynch, Michael A. Peshkin, J.E. Colgate,
“Kinematic Constraints for Assisted Single-Arm Manipulation”. International
Conference on Robotics and Automation (ICRA) 2002.
Peng Pan, Michael A. Peshkin, J.E. Colgate, Kevin M.
Lynch, “Static Single-Arm Force Generation with Kinematic Constraints”.
Journal of Neurophysiology, May 2005.
Peng Pan, Kevin M. Lynch, Michael A. Peshkin, J.E.
Colgate, “Human Interaction with Passive Assistive Robots”. IEEE 9th
International Conference on Rehabilitation Robotics, June 2005
ˇˇ
|
