Sensor interpretation and task-directed planning using perceptual equivalence classes

Consideration is given to how a robot may interpret its sensors and direct its actions so as to gain more information about the world, and to accomplish manipulation tasks. The focus is on general techniques for coping with uncertainty, specifically, to sense the state of the task, adapt to changes, and reason to select actions to gain information and achieve the goal. When the environment is integrated through sensors, one in effect views a projection of the world onto the space of possible sensor values. The structure of this sensor space and its relationship to the world is investigated. It is observed that sensors partition the world into perceptual equivalence classes that can serve as natural landmarks. By analyzing the properties of these equivalence classes a lattice and a bundle structure for the information available to the robot through sensing and action are developed. This yields a framework in which algorithms for sensor-based planning and reasoning are developed and characterized