Safe and Efficient Operation of Mobile Robots in Indoor Environments: A User-Centric Shared Control System with High-Level Navigation Capabilities

Hospitalization and isolation can be a traumatic experience for immunocompromised children, especially because they are separated from their families and friends. Social robots have been proposed as a way to improve the quality of care for children hospitalized in isolation by providing alternative means of social interaction and support. Remote control of such robots in a hospital setting, particularly where safety is a major concern, can be a daunting task for young patients. This dissertation introduces a multilevel shared control system for mobile robots, specifically companion robots in hospital-like indoor spaces. The system integrates user inputs with algorithmic semi-autonomous control at multiple levels of operation with the goal of only overriding user control to avoid collisions. At the foundational level, direct joystick control allows for immediate navigation, while higher levels introduce advanced functionalities such as corridor detection, corridor following, room-to-room navigation, and human-following capabilities. At the core of the system is User-Centric Tangent Bug for Blended Control, U-CenTB2, a safe and efficient blended control algorithm implementing a modified Tangent Bug with a risk assessment strategy. U-CenTB2 ensures safety through collision avoidance without prior knowledge of the environment. As a user operates the robot in a building, it dynamically recognizes corridors, adding on a corridor-follower mode that intelligently avoids obstacles and enhances remote operation convenience. The system's adaptability can also be extended to a human follower mode that allows it to follow a recognized person. Additionally, by constructing a topological map, it is able to conduct future high-level tasks such as autonomously returning home or navigating to specific rooms. To evaluate the performance of the algorithm, we present a simulation-based performance evaluation design for shared control algorithms by conducting batch simulations via Monte Carlo method. U-CenTB2 is evaluated through this methodology. The results demonstrate the algorithm’s efficacy in preventing collisions and adhering to user inputs, thereby offering a significant contribution to teleoperation of assistive mobile robots.