Frontiers in Robotics and AI最新文献

Exoskeleton assistance has the potential to address many gait related symptoms of Parkinson's disease (PD). However, gait variability, a hallmark of PD, makes designing exoskeleton controllers uniquely challenging. We sought to overcome the challenges that gait variability in PD poses for state estimation by employing machine-learning models for gait-phase estimation within our exoskeleton controller. Using machine-learning-based gait-phase models deployed on a hip exoskeleton (N = 7), we performed a 2-day protocol for people with PD where the first day focused on acclimation to the device and the second focused on evaluating the device by collecting gait metrics. Using 2-min walking tests, we assessed the impact of two different types of fixed torque assistance profiles on spatiotemporal and kinematic gait metrics. We demonstrated significant improvements to hip range-of-motion (8.4%), swing time (4.7%), and peak toe clearance (12.3%) in people with PD when walking with a combined flexion and extension assistance profile as compared to walking without an exoskeleton. Although we saw trends, there were no significant differences from providing only flexion assistance given our sample size. We also demonstrated that participant-specific models reduced gait-phase estimation error by 40%, however, resulting gait metrics were not significantly altered compared to metrics when walking with the generic model. These results demonstrate that ML gait-phase-based control approaches with limited PD-specific data can improve PD gait kinematics, with enhanced accuracy associated with participant-specific data. Ultimately, these results contribute to the goal of assistive exoskeletons in everyday use for people with Parkinson's disease.

Introduction: LiDAR-camera fusion systems are widely used in robotic localization and perception, where accurate extrinsic calibration is crucial for multi-sensor fusion. During long-term operation, extrinsic parameters can drift due to vibration and other disturbances, while target-based recalibration is inconvenient in the field and targetless approaches often suffer from highly non-convex objectives and limited robustness in challenging outdoor scenes.

Methods: We propose a targetless LiDAR-camera extrinsic calibration method by minimizing a semantic distribution consistency risk on SE(3). We align semantic probability distributions from the two sensing modalities in the image domain and freeze the pixel sampling measure at an anchor pose, so that pixel weighting no longer depends on the current extrinsic estimate and the objective landscape remains stable during optimization. On top of this anchor-fixed measure, we introduce a direction-aware weighting strategy that emphasizes pixels sensitive to yaw perturbations, improving the conditioning of rotation estimation. We further use a globally balanced Jensen-Shannon divergence to mitigate semantic class imbalance and enhance robustness.

Results: Experiments on the KITTI Odometry dataset show that the proposed method reliably converges from substantial initial perturbations and yields stable extrinsic estimates.

Discussion: The results indicate that the method is promising for maintaining long-term LiDAR-camera calibration in real-world robotic systems.

Introduction: Robotics and artificial intelligence (AI) are rapidly reshaping hospitality by automating frontline and back-of-house processes, augmenting service encounters, and expanding the analytical scope of revenue management. Yet, existing research remains fragmented: service-robot studies largely emphasize adoption and human-robot interaction, while revenue-management research prioritizes pricing and distribution, sustainability research focuses on environmental practices, and hotel real-estate scholarship foregrounds governance and asset value. Meanwhile, blockchain technologies-through distributed ledgers, smart contracts, digital identity, and tokenization-offer a complementary trust and value-transfer layer that can address coordination and verification problems across hotel ecosystems (e.g., data sharing, sustainability claims, and owner-operator contracting).

Methods: Drawing on an integrative literature synthesis, this conceptual article develops an integrative framework linking AI-robotics and blockchain capabilities to three interdependent hotel decision domains: (1) revenue management (demand forecasting, dynamic/open pricing, channel and loyalty optimization), (2) sustainability and operations (resource optimization, waste circularity, predictive maintenance), and (3) real estate and hotel asset management (digital twins, CapEx planning, valuation and risk analytics, and tokenized financing).

Results: A conceptual model is proposed in which AI-robotics and blockchain jointly build digital operational and market-intelligence capabilities that improve financial performance (RevPAR/GOPPAR and net operating income), sustainability performance (carbon and resource intensity), and long-term asset value. Ten propositions articulate mechanisms and boundary conditions related to governance, ethics, privacy, cybersecurity, organizational readiness, regulation, and market context.

Discussion: The article concludes with implications for hotel managers, owners, investors, and researchers, and outlines a future research agenda for hospitality, tourism, service management, and real-estate scholars.

Over the past decade, the field of social robotics has witnessed significant advancements in enhancing user experience (UX) and customer experience (CX) through the integration of Explainable Artificial Intelligence (XAI) and Interpretable Artificial Intelligence (IAI). This research presents a review that examines the progress made over a decade (2015-2025) in developing frameworks for social robotics and human-robot interaction (HRI) that prioritize transparency, trust, and user engagement. The journey began with early efforts to equip social robots with internal needs and motivations, forming the basis for understandable self-explanations. As the field progressed, there was a shift towards more user-centered approaches, autonomous social behavior, and self-explanations. By the early 2020s, researchers had begun to focus on the specific applications of XAI and IAI in social robotics. Past studies have shown that explainable and interpretable AI systems in social robots contributed to sustained user engagement and improved CX over extended periods. Currently, by 2025, the field has matured considerably, with researchers developing comprehensive frameworks that seamlessly integrated UX/CX considerations in social robotics with an emphasis on ethical considerations and societal implications. This research highlights how the past decade has seen remarkable progress in enhancing UX/CX in social robotics through XAI and IAI.

As introduced by Asimov in "I, Robot", intelligent machines are characterized as systems capable of performing tasks that traditionally require human intelligence, such as autonomous decision-making and driving. In this context, modern road vehicles can increasingly be understood as robotic systems endowed with progressively sophisticated functionalities, operational flexibility, and, crucially, the capacity to learn and evolve autonomously over time. Building on this perspective, AI-defined vehicles (AIDVs) are emerging in both the automotive industry and the research community as a next stage in vehicle evolution, where interaction capabilities, adaptability, sustainability, and ethical governance are embedded as core design principles rather than treated as auxiliary features. This work aims to introduce this new class of vehicles and provide an analysis of their defining principles, capabilities, and challenges. This article contributes a first conceptualization of AIDVs, outlines their defining principles, and distinguishes them from existing vehicle classes. Then, it identifies the risks introduced by adaptive AI and proposes a preliminary roadmap for their integration into Intelligent Transportation Systems (ITS).

Introduction: Social robots are increasingly being integrated into children's daily lives, shaping their social interactions and learning behaviors. However, no study has empirically investigated the effect of robot-administered praise in children younger than 4 years old.

Method: This study focuses on the social robot CommU, a simple, approximately 30 cm tall, child-shaped robot that exerts less social pressure and helps children attend to social cues more easily. We examined whether praise from CommU is associated with task persistence in children aged 18-24 months, in comparison with human praise.

Result: Children showed greater task persistence in the Praise condition than in the No Praise condition, regardless of agent type (CommU vs. Human). In addition, children's task persistence was positively associated with the amount of time they spent looking at the agent.

Discussion: These findings suggest that praise delivered by a social robot is associated with greater task persistence in children aged 18-24 months. Additionally, the positive association between task persistence and time spent looking at the agent suggests that children's social attention may contribute to sustained engagement during the task. More broadly, the results point to the possibility that social robots may be relevant to aspects of early childhood engagement, beyond the specific task-persistence behavior examined in this study.

Robotic systems are transforming image-guided interventions by enhancing accuracy and minimizing radiation exposure. A significant challenge in robotic assistance lies in surgical path planning, which often relies on the registration of intraoperative 2D images with preoperative 3D CT scans. This requirement can be burdensome and costly, particularly in procedures like vertebroplasty, where preoperative CT scans are not routinely performed. To address this issue, we introduce a differentiable rendering-based framework for 3D transpedicular path planning utilizing bi-planar 2D X-rays. Our method integrates differentiable rendering with a vertebral atlas generated through a Statistical Shape Model (SSM) and employs a learned similarity loss to refine the SSM shape and pose dynamically, independent of fixed imaging geometries. We evaluated our framework in two stages: first, through vertebral reconstruction from orthogonal X-rays for benchmarking, and second, via clinician-in-the-loop path planning using arbitrary-view X-rays. Our results indicate that our method outperformed a normalized cross-correlation baseline in reconstruction metrics (DICE: 0.75 vs. 0.65) and achieved comparable performance to the state-of-the-art model ReVerteR (DICE: 0.77), while maintaining generalization to arbitrary views. Success rates for bipedicular planning reached 82% with synthetic data and 75% with cadaver data, exceeding the 66% and 31% rates of a 2D-to-3D baseline, respectively. In conclusion, our framework demonstrates the feasibility of versatile, CT-free 3D path planning for robot-assisted vertebroplasty, accommodating diverse intraoperative imaging conditions without requiring preoperative CT scans.

There has been significant development in agricultural robotics over the past few years in the pursuit of optimising efficiency and addressing issues such as labour shortages and humans performing hazardous and arduous tasks. Despite this, human-robot interaction in the agricultural sector remains largely unchanged, often requiring technical expertise, which hinders wide-scale adoption. This problem is particularly pronounced in the African context, where limited technical exposure and linguistic diversity pose significant barriers to the adoption of these technologies. While alternative means for human-robot collaboration have been developed, these methods are currently limited to indoor structured environments. In this work, we introduce Osiris++, a flexible approach designed to allow seamless communication between robots and humans on an array of precision agriculture tasks. We validate and evaluate the performance of Osiris++ in real-world agricultural environments, demonstrating that the system can create accurate and useful scene graphs that aid in solving the assigned tasks. This paves the way for the possibility of allowing natural language instructions, including those in African languages, to be issued to robots within the agricultural sector.