Copyright The Robotics Lab 2008
Robots have recently graduated from structured laboratories to outdoor environments with varying and unstructured terrain. In order to be highly mobile and effective in these settings, robotics research will need to shift its focus to 1) understanding the underlying physics of robot/terrain interaction and 2) creating design methodologies for mechanically robust robots capable of multiple modes of locomotion such as running, leaping, and climbing. The Robotics Lab addresses these issues and, specifically, the following questions:
Robot/Terrain Interaction for Modeling, Control, and Behavior Development
Robot/Terrain Interaction for Modeling, Control, and Behavior Development
- How can vehicle/terrain interaction effects be exploited to improve control algorithms such that the robot is operating at the edge of its capabilities? For example, wheeled vehicles are able to execute vastly different turns on loose gravel compared to hard-packed soil. How can these distinctive behaviors be defined in terms of the terrain properties?
- How can a probabilistic approach to terrain modeling lead to more successful path planning? For instance, the trajectory space approach to hazard avoidance can account for terrain roughness and uncertainty by modifying its constraints. How can a similar approach be applied to more general path planning algorithms and behaviors?
- What principles will yield a systematic approach to designing and manufacturing robots for harsh environments from the initial design phase as opposed to hardening the robot as an afterthought? Can techniques such as Shape Deposition Manufacturing ensure robust operation?
- Multiple modes of locomotion require an increase in the number of degrees of freedom. How can we design robots that embody this functionality without drastically increasing complexity?
- How do we incorporate research being done in material science, the MEMS community, and other disciplines into novel uses for robots? What design, planning, and control strategies will take advantage of these materials and actuators?
RiSE at SwRI
