Human-Robot Collaboration in Cleaning Applications: Methods, Limitations, and Proposed Solutions

Abstract

As robotic technologies become more advanced, their integration into everyday tasks like cleaning becomes increasingly practical and necessary. Human-Robot Collaboration (HRC) bridges the gap between human ingenuity and robotic precision. Research in Human-Robot Collaboration for cleaning is essential to develop efficient, safe, and user-friendly robotic systems that can seamlessly integrate with human workflows in various cleaning environments. This study provides an in-depth analysis of the various models of Human-Robot Collaboration (HRC) in cleaning robots, their inherent limitations, and the proposed solutions to enhance their functionality and effectiveness. Cleaning robots exhibit diverse functionalities across different HRC models, such as Supervised Autonomy, Shared Control, Cobotic Systems, and several others. The discussed models feature the combinations of robotic abilities and human guidance, designed for the nature of the environment and the type of cleaning tasks. Every model demonstrates a specific cooperation between human operators and robots, with roles ranging from direct oversight to greater autonomy, where robots learn and adapt from human feedback. This study identifies five general limitations that are commonly associated with cleaning robots in HRC settings. These include limited flexibility in unstructured environments, difficulty handling complex tasks, dependency on human supervision, limited sensory perception, and challenges in effective human-robot interaction. To address these challenges, the paper proposes a series of existing technological solutions. These include the development of Sensory Fusion and Perception Algorithms, which integrate multispectral sensor arrays and perception algorithms for enhanced environmental mapping and obstacle recognition. Reinforcement Learning and Context-Aware AI Models are suggested to enable adaptive behavior and intelligent decision-making in dynamic environments. Real-Time Adaptive SLAM Techniques and Automated Surface Detection and Adaptation Systems are suggested to improve navigation and cleaning efficiency. The use of Natural Language Processing for Human-Robot Interaction, Robotic Manipulators with Enhanced Dexterity, Self-Monitoring and Predictive Maintenance Algorithms, and Robust Multi-Modal Human-Robot Interaction Frameworks are recommended.