Assisting 3D Indoor Positioning for Robot Navigation

Abstract

With the increasing employment of mobile robots to achieve different tasks in various applications, the need for localization and body position for these robots is increasing rapidly. Many techniques are proposed to calculate the precise coordinates of a robot based on the distances measured between the robot and a set of reference points. Also, internal sensors, such as accelerometers and gyroscopes, are used to detect the body position and the direction of the robot. However, the effect of obstacles in an indoor environment and sensor drifts still limit the applicability of such systems. Thus, in this study, a novel technique that uses one or more robots to compensate for the missing stationary points is proposed. The robots in the proposed technique collaborate to improve the positioning accuracy, by providing reference points to each other. Per each movement execution of one robot, the remaining robots remain stationary, to provide the required reference points. When the robot finishes the movement execution, its position is updated based on the signals collected from the other robots, in addition to the position calculated by the onboard sensors. Then, another robot is selected to execute its movement.. The results show that the proposed method has been able to improve the positioning accuracy, by increasing the number of collaborating robots, when the median function if used to select the coordinates of the robot, among the candidate positions.