|
|
Foundations of Robotics Seminar, April 26, 2010
Time
and Place | Seminar Abstract
ICRA 2010 Practice Talks
Effects of Nerve Signal Transmission Delay in Somatosensory Reflex Modeling Based on Inverse Dynamics and Optimization
Akihiko Murai
Disney Research
Optimizing Coordinate Choice for Locomoting Systems
Ross Hatton
Carnegie Mellon University - Robotics Institute
Hierarchical Planning Architectures for Mobile Manipulation Tasks in Indoor Environments
Ross Knepper
Carnegie Mellon University - Robotics Institute
NSH 1305
Talk 4:30 pm
Effects of Nerve Signal Transmission Delay in Somatosensory Reflex Modeling Based on Inverse Dynamics and Optimization
Human motion coordination is a long-standing research issue in biomechanics, and it should also have some implications for humanoid robot control. We have built a whole-body somatosensory reflex model based on our neuromusculoskeletal model and identified its parameters through non-invasive measurements and statistical analysis. Such models are crucial for analyzing and estimating signals in the nervous system. In this paper, we focus on signal transmission delay of the somatosensory reflex loop and investigate its relationship with the generalization capability of the reflex model. We obtain some sets of model parameters assuming different time delays using the data obtained from a stepping motion, and perform cross validations against stepping motions with different cycles as well as entirely different behaviors such as squat and jump. Interestingly, time delays close to the value expected from physiological properties show better cross validation results than others. This result suggests that relative simple reflex control can be generalized to multiple behaviors if the parameters are appropriate, and that robust control is possible even with large feedback delay.
Optimizing Coordinate Choice for Locomoting Systems
Animals often use gaits -- cyclic changes in shape producing a net displacement -- to move through their environments. In robotics, we are interested in planning motions for artificial systems that can match or exceed the locomotive capabilities of animals. A fundamental question of locomotion is 'What are the characteristics of a useful gait?' The geometric mechanics community has made significant progress in answering this question, identifying functions of the system shape that capture the net displacements induced by broad classes of gaits without having to individually test each possible motion. In this talk, we first introduce these results with the aim of separating them from the specialized language of differential geometry and making them accessible to a broader audience. Following this introduction, we then examine how the choice of generalized coordinates affects the quality of the locomotion functions, a question that has not previously been addressed.
Hierarchical Planning Architectures for Mobile Manipulation Tasks in Indoor Environments
This paper describes a hierarchical planner deployed on a mobile manipulation system. The main idea is a two-level hierarchy combining a global planner which provides rough guidance to a local planner. We place a premium on fast response, so the global planner achieves speed by using a very rough approximation of the robot kinematics, and the local planner begins execution of the next action even without considering subsequent actions in detail, instead relying on the guidance of the global planner. The system exhibits few planning delays, and yet is surprisingly effective at planning collision free motions. The system is deployed on HERB [20], combining a Segway mobile platform, a WAM arm, and a Barrett hand. The navigation and manipulation components have been tested on the real robot, and the task of simultaneously approaching and grasping a bottle on a countertop was demonstrated in simulation.
The Robotics Institute is part of the School of Computer Science, Carnegie Mellon University.