New localization strategy for mobile robot transportation in life science automation using StarGazer sensor, time series modeling and Kalman filter processing

A new mobile robot transportation system has been developed for modern life science laboratories. In the system, a new kind of sensors named StarGazer is adopted for the mobile robots´ indoor localization. To improve the positioning robustness of the StarGazer sensors under laboratory ceiling interference situations (such as, the strong ceiling lighting), a new hybrid signal filtering method is proposed by combing the TSM (Time Series Method) and the KF (Kalman Filter). In the proposed method, the KF algorithm is established to track and forecast the StarGazer based robot indoor positioning coordinates to avoid the mobile robots get lost in the interference environments and the TSM is adopted to choose the best initial parameters for the state and measurement equations of the KF model. A comparison of the proposed TSM-KF model and the standard ARIMA model is also provided in this paper. The results of an experiment show that the proposed hybrid method tracks and forecasts the robot indoor coordinates accurately, which can promote the robustness of the StarGazer based robot indoor navigation.