Publication

Abstract : Assistive mobility tools are gradually undergoing an evolutionary process to become Cyber-Physical Systems (CPS) that integrate perception, computation, and action for the safety of the user. This paper proposes an Internet of Things (IoT)-enabled smart walker that uses LiDAR-based Simultaneous Location And Mapping (SLAM) and safety-aware motion arbitration based on multi-sensor fusion and obstacle detection. This IoT-enabled smart walker integrates indoor mapping and obstacle detection achieved by the 360-degree LiDAR sensor, as well as ultrasonic, infrared, and inertial sensors. These various modes are fused for state estimation and obstacle identification, which are carried out by the Extended Kalman Filter. The computation component of the smart walker uses the Raspberry Pi, which processes the control component running at 10 Hz. An IoT component supported by the cloud provides for the monitoring of the caregiver, alerting for fall detection, and location information. Results show an accuracy of 4–7 cm for the 3-D maps, an average obstacle detection of 94.3%, and latency performance of less than 110 ms.