Keywords: Intelligent robots for rehabilitation, Medical and healthcare robotics
Abstract: Individualized and adaptive motor training is more effective than standard "one-fits-all" programs because it can account for individual needs, provide challenging tasks, and adapt to the user's performance. Automating this process is especially important in the face of a rapidly aging society and a shortage of skilled healthcare workers. The main challenges are automatic assessment of user performance, adaptive task difficulty, and dynamic exercise scheduling to provide an individualized training plan. The contributions of this paper are: 1) Overview of current literature on methods for individualized and adaptive motor training; 2) User study to evaluate the technical ability of the robotic walker RoboTrainer to provide individualized training; 3) Presentation of the novel framework RoboTrainerAID, which uses machine learning techniques for automatic assessment and spatial control actions for adaptive task difficulty. Evaluation of this framework will be conducted in a future user study. The goal of this paper is to highlight research gaps and provide concepts for the technical feasibility of robot-assisted, individualized, and adaptive lower-limb motor training and rehabilitation.

Keywords: Medical and healthcare robotics, Intelligent robots for rehabilitation, Socio-economic Impacts
Abstract: The ability of a prosthetic hand to mimic the biomechanics of a human hand is one of the approaches to assess its effectiveness. Despite significant advancements, replicating the dexterity and natural motion of a human hand remains a challenge. One critical yet often overlooked factor in prosthetic hand development is incorporating inter-phalangeal constraints and dimensional ratios, which play a key role in achieving natural finger articulation. To address these limitations, this study presents the kinematic and static analyses of ENRICH (EMG coNtrolled pRosthetIC Hand), a prosthetic hand designed with a biomimetic approach. The kinematic analysis evaluates joint variable relationships with actuation parameters, fingertip trajectories, and workspace analysis. The static analysis investigates the relationship between joint variables and actuator torque. Results indicate a linear correlation between joint variables and actuation parameters, while the estimated work-volume of ENRICH closely resembles 91.67% of a human hand’s work-volume. Additionally, the static analysis reveals that actuator torque follows a periodic function of joint variables. Due to the incorporation of inter-phalangeal constraints and dimensions, ENRICH demonstrates the ability to perform natural grasping.

Keywords: Service and assistive robotics, Intelligent robots for rehabilitation, Human Vehicle Interaction
Abstract: Wheelchairs, whether electric or manual, primarily serve to transport individuals, including those with disabilities or the elderly, from one place to another. Enhancing their quality of life has become a key societal challenge, driving the integration of robotic technologies into mobility solutions. Traditional robotic wheelchairs use a 2-DOF differential-drive platform, enabling linear motion and rotation but limited to transportation. This study explores expanding functionality by incorporating body movement control in electric wheelchairs. A 3-DOF mobility platform, Volting, introduces an additional movement, lateral tilt, enhancing mobility and adaptability. However, the comparative performance of 2-DOF and 3-DOF platforms remains unclear. To address this, an experimental study analyzed user experience during navigation and dancing tasks, chosen to represent conventional and exploratory movement. The study also included both visually normal and simulated visually impaired users (blindfolded participants). Results revealed distinct characteristics of each platform across different tasks and user groups, providing valuable insights for future robotic mobility design.

Keywords: HRI and social robotics, Medical and healthcare robotics, Service and assistive robotics
Abstract: Robots have the potential to impact society beyond traditional manufacturing and professional environments positively. Achieving this requires not only technical advancements but also addressing design challenges to tackle social issues effectively. One critical area where robots can make a substantial impact is in ensuring the equitable distribution of high-quality healthcare. However, existing robotic solutions are often costly and inaccessible, limiting their widespread adoption. To address this gap, this work presents Axobotl, a cost-effective and customizable social robot prototype designed to promote healthy habits among students and office workers. Axobotl leverages widely available components and modern hardware, enabling the integration of state-of-the-art AI capabilities. The prototype features a smartphone with a touch display, Dynamixel servo-actuated movements, and a Raspberry Pi-based control system, enabling interactive and expressive behaviors. This paper outlines the design and development process of Axobotl, which follows a human-centered design thinking and participatory design approach, including ideation, definition, and prototyping phases. Axobotl represents an initial step toward democratizing social robotics for preventive healthcare, particularly in developing countries.

Keywords: Intelligent robots for rehabilitation, Medical and healthcare robotics, Human Vehicle Interaction
Abstract: Stroke is a prevalent condition that often leads to significant cognitive and motor impairments. Traditional rehabilitation methods, while effective, are labor intensive and time consuming. In recent years, the integration of interactive games with rehabilitation devices has gained attention as a promising approach to improve patient motivation and engagement. This study evaluates the effectiveness of a hybrid rehabilitation model that combines a wrist rehabilitation device, using a bilateral wrist robot with compliant actuation, with interactive games for cognitive rehabilitation. The study compares the results of three rehabilitation models: traditional paper and pencil tasks, interactive game-based rehabilitation, and the proposed hybrid model. The results suggest that the hybrid model offers better cognitive rehabilitation, demonstrating better patient participation and improved cognitive function compared to traditional methods. This approach has the potential to optimize rehabilitation strategies and improve cognitive rehabilitation.

Keywords: Medical and healthcare robotics
Abstract: Effective hemorrhage control is critical in emergency medicine, as rapid and efficient intervention can significantly impact patient survival. However, traditional tourniquets that apply static pressure pose risks such as tissue damage, apoptosis, and thrombosis due to excessive compression. To address these risks, this work presents the design and fabrication of a soft robotic tourniquet capable of dynamically adjusting pressure to achieve Arterial Occlusion Pressure (AOP) while reducing tissue damage and clot formation. With soft-robotics ring inspired by the longitudinal muscle structures of tardigrades, the proposed design uses internal chambers that contract under negative pressure, applying circumferential compression around the limb. Through simulation-driven design refinement, iterative prototyping, and experimental testing on a venipuncture simulation arm, the tourniquet achieved an internal diameter reduction of 53.95%, demonstrating effective contraction and conformance to the limb. This paper details the research, design process, simulation, and experimental validation of the device, highlighting potential to improve hemorrhage control in medical emergencies. The results contribute to the advancement of wearable medical devices and outline future directions for improving tourniquets through soft robotics technology.

Keywords: Socio-economic Impacts, Service and assistive robotics, HRI and social robotics
Abstract: Cybernetic avatars (CAs) are key components of an avatar-symbiotic society, enabling individuals to overcome physical limitations through virtual agents and robotic assistants. While semi-autonomous CAs intermittently require human teleoperation and supervision, the deployment of fully autonomous CAs remains a challenge. This study evaluates public perception and potential social impacts of fully autonomous CAs for physical support in daily life. To this end, we conducted a large-scale demonstration and survey during Avatar Land, a 19-day public event in Osaka, Japan, where fully autonomous robotic CAs, alongside semi-autonomous CAs, performed daily object retrieval tasks. Specifically, we analyzed responses from 2,285 visitors who engaged with various CAs, including a subset of 333 participants who interacted with fully autonomous CAs and shared their perceptions and concerns through a survey questionnaire. The survey results indicate interest in CAs for physical support in daily life and at work. However, concerns were raised regarding task execution reliability. In contrast, cost and human-like interaction were not dominant concerns. Project page: https://lotfielhafi.github.io/FACA-Survey/.

Keywords: HRI and social robotics, The role of AI in the implementation of cognitive robots, Service and assistive robotics
Abstract: The hypothesis that a robot's active behavior improves people’s perception of agency toward the robot is examined through observation of interactions between human participants and a real robot and through questionnaires. Because of the cute appearance and childlike behavior of social robots, people have conventionally felt that the robots rely on people. In order for robots to be introduced into more situations, such as caregiving, it is also necessary for people to feel that they can rely on robots. The key to making people feel that they can rely on robots is to improve their perception of agency toward the robot. Humans are generally considered to have agency, and they behave actively based on their motivation to achieve certain goals. If a robot behaves actively based on the equivalent of a human's motivation, then the people's perception of agency toward the robot would be improved. We had human participants interact with a robot that had implemented the equivalent of a human's motivation using the Free Energy Principle. Although not statistically significant, the results suggest a possible trend toward improved perception of agency for robots that showed active behavior compared to those that did not.

Keywords: HRI and social robotics, Service and assistive robotics, Robot Vision and Sensing
Abstract: With the growing integration of autonomous mobile robots to everyday scenarios, encouraging natural and socially acceptable interactions between humans and robots becomes crucial. Despite significant advancements in artificial intelligence, many individuals remain hesitant to engage with robots because of a lack of technological knowledge and robot's inability to read social cues. This research addresses this gap by exploring how emotion recognition can enhance human-robot interaction and autonomous navigation, by allowing the robot to start an interaction if the approachability is positive, specifically, if the human facial expression denotes happiness. Experiments included both positive and negative approachability cases, with participants triggering corresponding detections. Data from three participants validated the approach. Our preliminary findings demonstrate that the system successfully enabled robots to approach humans when approachability was inferred as positive, enhancing interaction naturalness. These results highlight the potential of integrating emotion recognition into robotic systems, leading to socially intelligent autonomous robots.

Keywords: Humanoid Robots, HRI and social robotics
Abstract: Recently, there has been growing interest in human-robot interaction (HRI) for social robots. In particular, research has been conducted on generating HRI motions based on Large Language Models (LLMs) as one approach. This paper proposes a hierarchical motion control method to ensure the stable execution of HRI motions generated by LLMs. To achieve stable human-robot interaction, human-like movements must be generated while also considering multiple constraints to prevent damage to the robot. However, LLM-based motion generation faces challenges in fully accounting for these constraints when generating trajectories. To address this problem and realize stable HRI motion, this paper presents a hierarchical motion control method that ensures the execution of HRI motions while considering these constraints. To generate HRI motions on a dual-arm social robot, the end-effector trajectory of the robot is provided to the hierarchical motion controller. In the first layer of the hierarchical controller, the trajectory is converted into joint-level inputs while considering the robot’s structural constraints. Based on these target inputs for each joint, a variable optimization-based low-level controller is employed to execute the motions while satisfying the given constraints. The effectiveness of the proposed method is demonstrated through simulations.

Keywords: Entertainment robots, Humanoid Robots, HRI and social robotics
Abstract: We present a pneumatic soft robot designed to morphologically adapt to a wide range of human body shapes while minimizing actuator count to streamline both use and fabrication. A preliminary survey informed the robot’s adjustment parameters by identifying key user preferences and common fit challenges. The resulting prototype integrates ten inflatable cavities that modulate height, chest, waist, hips, and leg dimensions. To enable accessible and repeatable production, we developed a simplified mold-casting process using room-temperature vulcanizing silicone. We evaluated the system through a human-robot interaction study during a public open-space demonstration, collecting qualitative feedback on perceived adaptability and engagement. Results validate the effectiveness of our end-to-end workflow—linking user-informed design, fabrication, and testing—and highlight the robot’s potential for inclusive, interactive applications in retail and entertainment.

Keywords: HRI and social robotics, Traceability and Safety, Robotics for manufacturing
Abstract: Understanding the factors that influence perceived comfort in human-robot interaction (HRI) is critical for designing robotic systems that enhance user experience. This study investigates the effects of robot behavior and spatial layout on comfort, specifically whether these factors interact or contribute independently to user perception. Participants encountered a quadruped robot in a corridor under different behavioral conditions (moving, stopping, sitting) and corridor width (narrow, wide), and their comfort ratings were assessed using validated scales. Repeated measures ANOVAs revealed a significant main effect of robot behavior, with participants reporting greater comfort when the robot continued to move than when it stopped or sat. Additionally, a significant effect of aisle width was observed, with greater comfort reported in the wider aisle. However, there was no significant interaction effect, suggesting that these factors influence comfort independently. These findings highlight the importance of both robot behavior and environmental design in shaping positive user experiences in HRI. Future research should explore individual differences, such as attitudes toward quadruped robots, as well as additional environmental factors and related concepts to further optimize human-robot interactions.

Keywords: Legal and ethical regulations for AI and robotic systems, Socio-economic Impacts, HRI and social robotics
Abstract: Robots and AI systems are increasingly being developed for and deployed in contexts where decision-making entails moral implications. Examples include autonomous vehicles navigating the ethical dilemmas of traffic, healthcare robots tasked with ensuring patient safety and well-being, assistive robots upholding the autonomy and dignity of the elderly and people with disabilities, and social robots guiding children through educational experiences while assisting their emotional and cognitive development. The transition from passive tools to autonomous entities with moral decision-making capabilities has ignited extensive debate about the ethical responsibilities of Artificial Moral Agents (AMAs). This paper synthesizes recent advancements in AMA research, addressing evolving debates on their ethical feasibility and societal integration. We assess key arguments against and for AMAs, highlighting impacts on moral responsibility, cultural perspectives, and stakeholder trust. Our analysis reveals that while AMAs remain a subject of theoretical debate, their integration into ethically sensitive contexts is increasingly proposed, necessitating clearer governance strategies. Given the recent AI advancements and increased deployment of robotics in high-stakes settings, this synthesis is timely and highlights the urgency of addressing these ethical challenges. By analyzing recent advancements and diverse perspectives, we aim to provide a concise but comprehensive understanding of the complexities involved in empowering machines with moral decision-making capabilities.

Keywords: Robotic education and robots for education, Robot Vision and Sensing, Education and Skill Transfer
Abstract: Educational Robotics (ER) focuses on developing tools to enhance learning experiences. Simulators are recurrent in ER due to their convenience as substitutes for physical setups, providing assistance when teaching real robotics applications. However, one limitation is often the achievable level of realism. At Tecnologico de Monterrey, students majoring in robotics learn about autonomous robots through integration projects during their junior and senior years. To expand the availability of the required test courses, we propose a scene generation pipeline to create affordable simulated environments. We developed two versions of a simulator featuring a photorealistic 3DGS environment, LiDAR signals, and a ROS2 interface. We validated our proposal through trial runs with both real and simulated robots, as well as an assessment of image quality.

Keywords:
Abstract: AI systems and robots are commonly deployed in various environments in which they interact with people. This led to increasing humanization of design and capabilities for easier integration and to make interaction with users more intuitive, including humanoid shape and behaviour, use of natural language, or emotion detection and synthesis.  But what is the social status of such AI systems and robots? Are they mere tools? Assistants? Collaborators? Companions? A new species? Scientific and technological foundations of Robotics and AI, and studies about human behavior must be recalled to help shed some light on this ongoing debate. The way we decide to consider machines will shape our future societies.

Keywords: Robot Vision and Sensing, Machine Learning and Robot Learning
Abstract: Tactile sensors are vital for robotic manipulation, providing critical data on contact force magnitude and distribution to detect grasping states. However, their integration with rigid grippers often limits adaptability and efficacy, particularly when handling delicate objects. Similarly, while soft electroadhesive grippers offer advantages in grasping fragile materials, challenges in deformation, compactness, and precise control persist due to inadequate sensing capabilities. Conventional combinations of tactile sensing and electroadhesion gripping, which typically relies on multilayer fabrication, usually result in hardness mismatches that impair performance. To overcome these limitations, this study presents a novel soft gripper that seamlessly integrates tactile sensing and electroadhesive functionalities within a monolithic silicone membrane. This design enables precise force detection, adaptability to diverse object geometries, and reliable gripping. Experimental results demonstrate effective manipulation of small conductive and insulating objects, enhancing robotic efficiency and safety.

Keywords: Machine Learning and Robot Learning, Robotics for manufacturing, Industrial robotics
Abstract: This paper addresses the challenges of applying deep reinforcement learning (RL) to robotic grasping tasks, focusing on bridging the sim-to-real gap to ensure seamless transfer from simulated to real-world environments. We propose a two-phase modeling approach: (1) an image-free model designed to validate core functionalities and (2) an image-based model that incorporates visual data to enhance grasping precision. Using QT-Opt as a foundation, the models are trained to learn optimal grasping actions based on state-action pairs in both simulation and real settings. To stabilize training and improve learning efficiency, we employ multiple neural networks enhanced with Polyak averaging and clipped double Q-learning to address gradient instability. For action optimization in continuous control tasks, the Cross Entropy Method (CEM) is integrated to ensure robust policy learning. Additionally, we leverage the Jump-Start Reinforcement Learning (JSRL) method to improve convergence, providing initial guidance from a non-image-based agent to enhance exploration efficiency. Experimental results show the effectiveness of different reward functions and the impact of camera positioning on model performance. We also explore the use of generative frameworks to mitigate visual discrepancies between simulated and real images, improving robustness in real-world grasping tasks. We conducted comprehensive experiments to evaluate the effects of various reward functions and the influence of camera positioning on model performance.

Keywords: Machine Learning and Robot Learning, Industrial robotics
Abstract: This paper presents a novel method that utilizes Wasserstein Generative Adversarial Networks (WGANs) with Gradient Penalty (GP) to approximate the distribution of the free conditioned configuration space. Our proposed approach involves conditioning the WGAN-GP with a Variational Auto-Encoder in a continuous latent space to handle multimodal datasets. However, training a Variational Auto-Encoder with WGAN-GP can be challenging for image-to-configuration-space problems, as the Kullback-Leibler loss function often converges to a random distribution. To overcome this issue, we simplify the configuration space as a set of Gaussian distributions and divide the dataset into several local models. This enables us to not only learn the model but also speed up its convergence. Our experiments show promising results for approximating the configuration space of unknown scenes. The proposed method is a step toward our goal of accelerating path planning tasks in unknown scenes with obstacles.

Keywords: Machine Learning and Robot Learning
Abstract: Soft robots can be unintuitive and difficult to design due to the complex and non-linear behavior of the materials used. Many works have relied on human intuition for their design, but this process is not ideal and is also labor intensive, so numerous techniques have been devised to assist in designing these soft robots. In recent times, Deep Reinforcement Learning (DRL) has been adapted and utilized as a black-box optimizer, where it has shown promising results in fields such as fluid mechanics, but it has not been explored in the field of soft robotics. In this paper we propose a DRL-based black-box optimizer utilizing the Deep Deterministic Policy Gradient (DDPG) algorithm for the optimization of a soft robotic finger. A reward function is formulated with the design goal of increasing the compliance of the soft robotic finger over a baseline design while minimizing the trade-off to the maximum stress experienced. The results demonstrate that the proposed approach is able to thoroughly explore the design space and generate suitable designs with consideration to the design goals.

Keywords: Industrial robotics, Robots in the Smart Factory, Service and assistive robotics
Abstract: Self-driving laboratories are automated research environments that utilize advanced technologies to conduct experiments and analyze data with minimal human intervention. The presence of robotic manipulators within an autonomous driving research facility plays a crucial role in carrying out tasks effectively. Motion planning schemes are used for planning motions of a manipulator equipped inside self-driving laboratories to facilitate the transition from an initial pose to a final pose during a task execution. In this paper, three motion planning schemes developed based on Jacobian methods are implemented to traverse a redundant manipulator with a coupled finger gripper through given trajectories. RRT* algorithm is used for planning trajectories and inverse solutions of the manipulator are computed separately using three Jacobian based methods such as Jacobian Transpose (JT), Pseudo Inverse (PI), and Damped Least Square (DLS) methods. Smoothness and RMSE pose errors of end-effector motions along with velocity continuity, acceleration profile, jerk, and snap values of joint motions are analysed for determining an efficient motion planning method. Advantages and disadvantages of the proposed motion planning schemes mentioned above are evaluated using simulation studies to determine a suitable inverse solution technique for the tasks. This paper demonstrates motion planning algorithms that achieve precise positioning while ensuring the efficient operation of robotic manipulators operating in service sectors, characterized by smoother trajectories, reduced computational, and energy demands.

Keywords: Intelligent Welding Robots and Systems, Machine Learning and Robot Learning, Robotics for manufacturing
Abstract: Achieving consistent performance, and high-quality brazed joints for copper tubes poses significant challenges due to the intricate interplay of thermal dynamics, material properties, and robot motion control. This paper presents a reinforcement learning (RL)--based framework for optimizing robotic trajectories and parameters in manufacturing brazing applications. The learned policy successfully guided the end-effector near the target region, as verified by the final pose.

Keywords: Machine Learning and Robot Learning, Intelligent Welding Robots and Systems, Robotics for manufacturing
Abstract: Over the past few decades, tensegrity structures have garnered significant interest among robotics researchers due to their high payload capacity, lightweight construction, and resistance to external shocks. Despite these advantages, certain aspects, such as the impact of the number of wires and rods on the robot's stiffness, remain underexplored. This study aims to fill this gap through experimental analysis, focusing on the robustness of tensegrity structures against external forces influenced by the number of wires and struts. We present an experimental comparison of operational consistency under various conditions, including payload, no payload, and vibration tests, between two prototypes: one with a triangular prismatic shape and another with a quadrangular shape featuring a larger number of tensegrity elements. The primary objective of this research is to evaluate the structural stability and payload capacity of both prototypes based on the number of wires and bars.

Keywords: Service and assistive robotics
Abstract: Autonomous robots have reached the physical capabilities to perform useful tasks in mixed-initiative human-robot teams. In this paper we provide an overview of the latest version of our DIARC architecture and show how it can be used in practical setting to quickly instruct a group of robots to perform novel tasks, in this instance food preparation tasks. Specifically, we motivate essential architectural capabilities required for effective mixed-initiative human-machine teaming and briefly discuss how DIARC meets them. We then introduce as an example of practical application a multi-robot, multi-human food assembly task as it occurs in fast food restaurants and show how different aspects of the task can be quickly taught through natural language dialogues, with the system immediately able to perform those tasks.

Keywords:

Keywords:

Keywords: Legal and ethical regulations for AI and robotic systems, HRI and social robotics, Legal Aspects
Abstract: The integration of robots into daily life raises complex ethical challenges, requiring a shift beyond traditional engineering norms. This paper applies Michel Foucault’s “Aesthetics of Existence” to robotics ethics, advocating for a framework where engineers engage in continuous self-reflection to shape human-robot cohabitation responsibly. To operationalize this approach, we propose applying agile governance, a dynamic regulatory model that adapts through multi-stakeholder collaboration and real-time ethical evaluation. Unlike static regulations, agile governance evolves alongside technological and societal changes, ensuring that robotic systems remain ethically aligned and socially responsible. We illustrate this framework through two case studies. PARO, a therapeutic robot for dementia care, demonstrates both its benefits and the unintended risk of social isolation, highlighting the need for ethically guided design and policy interventions. NICOBO, a companion robot under study in the JST RISTEX RInCA program, provides insights into the long-term governance of emotional human-robot relationships. These cases underscore the importance of iterative governance structures that integrate ethics, engineering, and user experience. Through these examples, we assess how governance models can be refined to ensure that robots enhance rather than diminish human sociality. This work provides both theoretical insights and practical guidelines, contributing to a future where human-robot cohabitation is ethically adaptive and socially transformative.

Keywords: HRI and social robotics
Abstract: This study extends the concept of Human-Object Interaction (HOI) to Human-Robot Interaction (HRI), examining how care (attentive handling of robots) and learning (gaining insights through interactions) relate to users’ well-being. A survey of 397 robot users in Japan analyzed the relationships between usage frequency, care, learning, loneliness, and life satisfaction. Results showed that more frequent use was associated with greater care and learning. Care was linked to lower loneliness, while learning was associated with higher life satisfaction, suggesting distinct psychological effects. Although robot type did not affect care or well-being, learning was higher for communicative and mobile robots than for cleaning robots. However, the study did not fully address differences among robot types or clarify distinctions between robots and static objects. Moreover, participants were limited to current robot owners, possibly biasing results. Future research should examine how robot characteristics influence psychological processes and expand to broader, more diverse populations to better understand human-robot coexistence.

Keywords: Legal and ethical regulations for AI and robotic systems, HRI and social robotics, Entertainment robots
Abstract: This study examines how social robots foster companionship through “yorisoi” (emotional closeness). Unlike conventional perspectives in HRI that aim to facilitate the practical application of robotics by fostering seamless human-robot interactions, this research highlights the role of unpredictability and misalignment in shaping dynamic relationships. Through qualitative analysis of Panasonic’s NICOBO owners’ experiences from an anthropological point of view, the study reveals that the robot’s imperfect responses and communication gaps encourage emotional engagement, prompting users to reinterpret interactions, adjust their expectations, and integrate the robot into their daily routines. Introducing the concept of “polyphonic companionship,” this study argues that human-robot bonds do not necessarily emerge from perfect harmony but through continuous negotiation and adaptation. Rather than merely simulating human-like companionship, NICOBO’s design fosters a fluid and evolving relationship, where meaning is co-created through ongoing interaction. This perspective challenges prevailing design paradigms that prioritize predictability, offering new insights into the nature of human-robot companionship.

Keywords: