Smartphone IMU-driven human-following robot with ESP32 threshold navigation
Abstract
This paper presents a cost-effective human-following robot utilizing ESP32 microcontroller and smartphone IMU sensor integration via Bluetooth communication.[1] The system employs accelerometer-based step detection and compass-derived azimuth data, eliminating the need for expensive LiDAR or thermal imaging sensors. Key technical contributions include a 30° azimuth threshold turning algorithm with step-synchronized forward movement, enabling reliable person-following within 10-meter range. Implementation demonstrates 6+ hours battery life using standard USB power bank and achieves 95% following accuracy across straight paths and turns. Targeting elderly assistance and warehouse applications, our pragmatic design leverages ubiquitous smartphone sensors with complete hardware-software validation. The paper details system architecture, navigation algorithms, performance testing, and practical deployment considerations for assistive robotics applications. Keywords: human-following robot, ESP32, Bluetooth navigation, assistive robotics, smartphone IMU sensors, azimuth, human gait Received: 21 March 2026 Accepted: 12 May 2026 Published: 21 May 2026
