Pathfinding Decision-Making Using Proximity Sensors, Depth Camera and Active IR Marker Tracking Data Fusion for Human Following Companion Robot

Assistive companion robots face a multitude of problems presented by a combination of environmental and hardware factors in addition to the nature of their operation. This is most evident in robots developed for following human subjects in indoor locale without the aid of an embedded environment or complex navigational facilities such as Simultaneous Localization and mapping (SLAM). This paper centers on the implementation and initial simulation of a navigation system that is developed for CARMI, a Companion Avatar Robot for the Mitigation of Injuries. This system utilizes the Microsoft Kinect's depth camera, an Active IR Marker tracking system and the robot's ultrasonic perimeter sensor array for simultaneous human following and obstacle avoidance while centering the Kinect device in the direction of an autistic child. Initial simulation results are presented to show successful functional capabilities of the system at discerning an approach direction based on depth, ranging and marker sensor data fusion. The conclusion of this paper greenlights simulations for more extensive navigational scenarios.