We propose the Planning of Landmark Sensing (PLAS) considering environmental conditions for autonomous mobile robots. When an autonomous mobile robot moves, the robot usually resets the accumulated errors by sensing the landmark. In the previous work, the errors of robots are estimated based on the robot model, and Kalman Filter is used to reset the accumulated error after the landmark sensing. Based on this method, only the relation of robot and landmark position is considered as observation noise. But in actual visual landmark sensing, environmental conditions, like position of lighting, brightness of room and so on, are likely to cause the misrecognition of landmark. Then we consider environmental conditions for the calculation of the observation noise of Kalman Filter. This helps our system to determine the possibility of the misrecognition in case of bad sensing environments. Further, we also calculate sensing reliability based on environmental conditions, and use this sensing reliability in the adaptation of wheel noise of the robot model. As a result of the consideration of environmental conditions, our navigation system can navigate robots precisely, even if the environments are not suitable for landmark sensing.
