Lightweight Two-Layer Control Architecture for Human-Following Robot.

(1) Background: Human detection and tracking are critical tasks for assistive autonomous robots, particularly in ensuring safe and efficient human-robot interaction in indoor environments. The increasing need for personal assistance among the elderly and people with disabilities has led to the development of innovative robotic systems. (2) Methods: This research presents a lightweight two-layer control architecture for a human-following robot, integrating a fuzzy behavior-based control system with low-level embedded controllers. The system uses an RGB-D sensor to capture distance and angular data, processed by a fuzzy controller to generate speed set-points for the robot's motors. The low-level control layer was developed using pole placement and internal model control (IMC) methods. (3) Results: Experimental validation demonstrated that the proposed architecture enables the robot to follow a person in real time, maintaining the predefined following distance of 1.3 m in each of the five conducted trials. The IMC-based controller demonstrated superior performance compared to the pole placement controller across all evaluated metrics. (4) Conclusions: The proposed control architecture effectively addresses the challenges of human-following in indoor environments, offering a robust, real-time solution suitable for assistive robotics with limited computational resources. The system's modularity and scalability make it a promising approach for future developments in personal assistance robotics.