Using BIM and Sensing Mats to Improve IMU-based Indoor Positioning Accuracy

Currently, numerous approaches to Indoor Positioning Systems (IPSs), such as RSSI (Received Signal Strength Indication), fingerprint, PDR (Pedestrian Dead-Reckoning), and image recognition, have been developed. But each individual positioning method has unique drawbacks. In this study, we provide an IPS with a novel combined positioning method that applies Building Information Modelling (BIM) and Internet of Things (IoT). We employ an Inertial Measurement Unit (IMU) to track people’s positions. We then utilize a BIM model that has information (semantic and geometric) and a sensing mat to eliminate IMU drift error in the positioning process. The demonstration field is a research office, and test results show that the BIM based positioning constraint can effectively filter IMU cumulative error along with time; thereby, positioning accuracy can be controlled to a range of 30cm × 30cm. In sum, this paper proposes a new positioning method that compensates for the weakness of the IMU. In the future, this system can be applied to people management, such as telecare for older adults.