AN AUGMENTED STEADY HAND SYSTEM FOR PRECISE MICROMANIPULATION

Steady Hand cooperative manipulation is a hands-on
approach that integrates seamlessly in the surgical
practice. In steady hand manipulation, the tool is
held simultaneously by the user and the robot and
the robot complies to forces applied by the user.
Steady hand manipulation promises significant
improvements in safety, accuracy over conventional
practice at minimal cost and training to the user. It
also offers a way around the difficult problem of
encoding human intelligence, and preserves the
benefits from experience and training.
We explore the possibility of encoding/utilizing task
descriptions to improve transparency and performance
of a steady hand manipulation task. This is done by
constructing higher level representations of the
task. The user''s interaction with the robot,
tool-tissue interactions, and other sensory and
planning inputs can be used to identify the task
state and modify the behavior of the robot by using
using optimized task and control parameters.
Validation experiments for several cooperative tasks
with and without augmentation are presented.

Autorentext
Rajesh Kumar received a PhD in Computer Science from the Johns Hopkins University in 2001. Before rejoining the research faculty of the university in 2007, he worked at Foster-Miller, Inc. and Intuitive Surgical, Inc. His research interests are in medical applications of robotics and computing.

Klappentext
Steady Hand cooperative manipulation is a hands-on approach that integrates seamlessly in the surgical practice. In steady hand manipulation, the tool is held simultaneously by the user and the robot and the robot complies to forces applied by the user. Steady hand manipulation promises significant improvements in safety, accuracy over conventional practice at minimal cost and training to the user. It also offers a way around the difficult problem of encoding human intelligence, and preserves the benefits from experience and training. We explore the possibility of encoding/utilizing task descriptions to improve transparency and performance of a steady hand manipulation task. This is done by constructing higher level representations of the task. The user's interaction with the robot, tool-tissue interactions, and other sensory and planning inputs can be used to identify the task state and modify the behavior of the robot by using using optimized task and control parameters. Validation experiments for several cooperative tasks with and without augmentation are presented.