Toward an occupancy model rendering framework for robotic systems operating under human supervision

The introduction of 3D modeling with probabilistic encoding of occupancy state of space, lead to the development of powerful tools to generate rich representations of environments. But the sophistication of the modeling scheme introduces important concerns in the way such representations can be efficiently displayed to human operators for validation purpose and decision making in the guidance of robotic systems. This paper proposes a method and describes the characteristics of the prototype that has been designed for probabilistic occupancy grid rendering with the objective to alleviate the relative complexity associated with the visual interpretation of 3D probabilistic occupancy grids. Various aspects of the interaction with the operator are investigated and suitable approaches are identified to provide maximum flexibility. Examples are presented of an application of the proposed rendering framework in telemanipulation operations.