An Information Theoretical Approach to View Planning with Kinematic and Geometric Constraints

We consider the view planning problem where the sensor, a range scanner, is mounted on a robot mechanism with non-trivial geometry and kinematics. The robot+sensor system is required to explore the environment (obstacle/free space). We present a novel information theoretical approach in which the sensing action is viewed as reducing ignorance of the planning space, the C-space of the robot. The concept of C-space entropy is introduced as a measure of this ignorance. The next view in the planning process is determined by maximizing the expected reduction of C-space entropy, called maximal entropy reduction (MER) criterion. A computational tool to implement MER is the notion of information gain density function (IGDF). Experimental results with a real PUMA robot with a wrist-mounted range scanner and a simulated robot show the effectiveness of the MER criterion in efficient exploration of environments for motion planning problems.