Collaborative and companion robots are at the forefront of technological innovation, transforming human-robot interaction to address a wide range of tasks and activities. This review provides a comprehensive examination of the current state of research on human following and guidance with autonomous mobile robots. Covering the evolution of research from the inception to the latest advancements in this area, we categorize existing literature based on various attributes including fields of application, technologies employed, and social acceptability. We critically analyze and compare state-of-the-art approaches in perception, tracking, planning, control, and human-robot interaction, highlighting their effectiveness and feasibility. We further classify studies based on application domains where person following and guiding tasks are particularly impactful, such as healthcare, personal assistance, logistics, and tour guiding. We identify persistent challenges and outline open problems, offering recommendations for future research directions. Our review aims to serve as a foundational reference for researchers and practitioners, fostering continued innovation and development in the deployment of autonomous robots for human following and guidance.
Population aging and pandemics have been shown to cause the isolation of elderly people in their houses, generating the need for a reliable assistive figure. Robotic assistants are the new frontier of innovation for domestic welfare, and elderly monitoring is one of the services a robot can handle for collective well-being. Despite these emerging needs, in the actual landscape of robotic assistants, there are no platforms that successfully combine reliable mobility in cluttered domestic spaces with lightweight and offline Artificial Intelligence (AI) solutions for perception and interaction. In this work, we present Marvin, a novel assistive robotic platform we developed with a modular layer-based architecture, merging a flexible mechanical design with cutting-edge AI for perception and vocal control. We focus the design of Marvin on three target service functions: monitoring of elderly and reduced-mobility subjects, remote presence and connectivity, and night assistance. Compared to previous works, we propose a tiny omnidirectional platform, which enables agile mobility and effective obstacle avoidance. Moreover, we design a controllable positioning device, which easily allows the user to access the interface for connectivity and extends the visual range of the camera sensor. Nonetheless, we delicately consider the privacy issues arising from private data collection on cloud services, a critical aspect of commercial AI-based assistants. To this end, we demonstrate how lightweight deep learning solutions for visual perception and vocal command can be adopted, completely running offline on the embedded hardware of the robot.
Robot assistants and service robots are rapidly spreading out as cutting-edge automation solutions to support people in their everyday life in workplaces, health centers, and domestic environments. Moreover, the COVID-19 pandemic drastically increased the need for service technology to help medical personnel in critical conditions in hospitals and domestic scenarios. The first requirement for an assistive robot is to navigate and follow the user in dynamic environments in complete autonomy. However, these advanced multitask behaviors require flexible mobility of the platform to accurately avoid obstacles in cluttered spaces while tracking the user. This paper presents a novel human-centered navigation system that successfully combines a real-time visual perception system with the mobility advantages provided by an omnidirectional robotic platform to precisely adjust the robot orientation and monitor a person while navigating. Our extensive experimentation conducted in a representative indoor scenario demonstrates that our solution offers efficient and safe motion planning for person-following and, more generally, for human-centered navigation tasks.
Learning agents can optimize standard autonomous navigation improving flexibility, efficiency, and computational cost of the system by adopting a wide variety of approaches. This work introduces the PIC4rl-gym, a fundamental modular framework to enhance navigation and learning research by mixing ROS2 and Gazebo, the standard tools of the robotics community, with Deep Reinforcement Learning (DRL). The paper describes the whole structure of the PIC4rl-gym, which fully integrates DRL agent's training and testing in several indoor and outdoor navigation scenarios and tasks. A modular approach is adopted to easily customize the simulation by selecting new platforms, sensors, or models. We demonstrate the potential of our novel gym by benchmarking the resulting policies, trained for different navigation tasks, with a complete set of metrics.
Population ageing and pandemics recently demonstrate to cause isolation of elderly people in their houses, generating the need for a reliable assistive figure. Robotic assistants are the new frontier of innovation for domestic welfare, and elderly monitoring is one of the services a robot can handle for collective well-being. Despite these emerging needs, in the actual landscape of robotic assistants there are no platform which successfully combines a reliable mobility in cluttered domestic spaces, with lightweight and offline Artificial Intelligence (AI) solutions for perception and interaction. In this work, we present Marvin, a novel assistive robotic platform we developed with a modular layer-based architecture, merging a flexible mechanical design with cutting-edge AI for perception and vocal control. We focus the design of Marvin on three target service functions: monitoring of elderly and reduced-mobility subjects, remote presence and connectivity, and night assistance. Compared to previous works, we propose a tiny omnidirectional platform, which enables agile mobility and effective obstacle avoidance. Moreover, we design a controllable positioning device, which easily allows the user to access the interface for connectivity and extends the visual range of the camera sensor. Nonetheless, we delicately consider the privacy issues arising from private data collection on cloud services, a critical aspect of commercial AI-based assistants. To this end, we demonstrate how lightweight deep learning solutions for visual perception and vocal command can be adopted, completely running offline on the embedded hardware of the robot.
5 March 2026 in International Journal of Social Robotics
Recent advancements in indoor robot navigation have driven the increasing presence of service robots in homes and public spaces. While traditional geometric planning methods excel at obstacle avoidance, they often fail to generate human-aware trajectories, limiting their effectiveness in social environments. Social navigation aims to bridge this gap by prioritizing socially acceptable behaviours, ensuring natural human-robot coexistence. Current research predominantly focuses on local control approaches that predict human motion to enhance safety. However, in crowded environments, robots must exhibit human-like behaviour that cannot be effectively addressed through reactive control alone. An integration at the planning level is deemed necessary to prevent disruptions to social dynamics. This study introduces a novel framework that enriches robotic navigation with adaptive social considerations, specifically targeting scenarios where local control methods fall short in maintaining social compliance. Our approach builds upon the robustness of classical grid-based planners while incorporating a learning-based social cost layer. Using an encoder-decoder neural model, we generate a dynamic social cost map from positional data of individuals, environmental geometry, and the robot’s goal. This cost map and the static obstacle map are integrated into the planning process, allowing the robot to navigate complex social settings without compromising the default planner functionality. We validate our method across diverse real-world and simulated scenarios, including queuing, group conversations, narrow blind passages, and corridor navigation. The results demonstrate the framework’s adaptability, robustness, and ability to generalize to different social contexts, ensuring socially aware robot navigation in dynamic environments.