Frontiers in Robotics and AI最新文献

{"title":"Structural predictors and latent maturity regimes of robotic readiness in global health systems: evidence from machine learning-based latent clustering and class prediction.","authors":"Moumita Mukherjee, Raja Hashim Ali","doi":"10.3389/frobt.2026.1835473","DOIUrl":"10.3389/frobt.2026.1835473","url":null,"abstract":"<p><strong>Background: </strong>The systematic integration of robotics into health service delivery systems requires periodic assessment of robotic readiness in terms of digital-health maturity regimes across countries. The current study aims to cluster 169 countries into maturity regimes and classify and predict cluster membership accuracy based on digital-health maturity dimensions determining the system's perception and interoperability, coordination, and workforce-regulatory reliability readiness. These country-level proxy concepts are applied due to a dearth of cross-country robotic readiness measures at the global level. The study also proposed an adaptive readiness framework for robotic deployment decision-making.</p><p><strong>Methods: </strong>This study applies a multilayered robotic systems readiness proxy framework based on diffusion of innovations and multi-agent systems theories, which hypothesise health-system readiness as adaptive and learning-driven instead of static. Using 31 global digital-health maturity indicators from the Global Digital Health Monitor 2023, three latent proxy dimensions are constructed to measure perception/interoperability readiness, governance/coordination readiness, and workforce-regulatory reliability readiness. The data represent 169 countries at different stages of digital-health maturity, starting from phase 1 to phase 5, where 67 countries reflect evidence-based digital-health maturity status and 102 countries had no data reported and are below phase 1. The sample included these 102 countries in the study, imputing the lowest values for each indicator to prioritise them and eliminate participation bias in global health policy decisions. To address concerns regarding missing data and its handling, the analysis was repeated across different robust scenarios. The latent proxy dimensions are created using principal component analysis, reducing 31 dimensions to 3 dimensions. Their associations were explored using ordinary least squares regression. Unsupervised clustering techniques (K-means, agglomerative hierarchical clustering, and Gaussian mixture modelling) were applied to identify latent readiness regimes, and comparative model assessment brought a four-cluster solution. External validation was conducted against the World Bank classification of countries. The separability of the latent regimes was assessed through repeated cross-validations and post-clustering recoverability analysis using conservative supervised learning models, where five latent clusters were considered in line with the World Health Organization's five-phase maturity categories.</p><p><strong>Results: </strong>The three proxy indices have shown internally coherent loading structures with 85%-95% scale reliability. Significant positive structural associations are evident where the perception/interoperability (β = 0.343, p < 0.01; observed-only scenario) and governance/coordination readiness (β = 0.838, p < 0.01; baseline scenario) l","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1835473"},"PeriodicalIF":3.7,"publicationDate":"2026-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13419970/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148632247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Benchmark evaluation in task and motion planning using iteratively deepened AND/OR graph networks.","authors":"Hossein Karami, Antony Thomas, Fulvio Mastrogiovanni","doi":"10.3389/frobt.2026.1853007","DOIUrl":"10.3389/frobt.2026.1853007","url":null,"abstract":"<p><p>In robotics research, each subdomain presents a distinct set of challenges, and any framework designed for a given domain must effectively address these complexities. However, a single application within that domain may not fully capture the breadth of challenges inherent to it. To enable systematic and comprehensive evaluation, the robotics community has developed standardized problem scenarios and associated performance metrics, commonly referred to as benchmarks, which collectively represent the diverse challenges arising across applications. In this work, we evaluate our task-motion Planning (TMP) framework on five benchmarks proposed by the TMP community. We begin by briefly describing our iterative deepening AND/OR graph-based TMP planner. Subsequently, we assess its performance across these benchmarks, each designed to capture different aspects of the challenges in TMP. The evaluation demonstrates that the proposed planner successfully solves all five benchmarks, thereby indicating that our framework constitutes a robust and effective solution for TMP.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1853007"},"PeriodicalIF":3.7,"publicationDate":"2026-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13414962/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148621859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Object-centric diffusion policies for real-world robotic-arm imitation learning.","authors":"Prashant Reddy Kasu, Dugan Um","doi":"10.3389/frobt.2026.1829436","DOIUrl":"10.3389/frobt.2026.1829436","url":null,"abstract":"<p><p>Imitation learning in complex, unstructured environments remains challenging due to the difficulty of grounding perception in physically meaningful representations and the need to model multimodal action distributions. Existing approaches often rely on unstructured pixel-level feature encodings or stochastic latent-variable decoders, which can lead to brittle attention in cluttered scenes. In this work, we present a novel integration of detector-based visual representations with conditional diffusion modeling (DINO + CDP) for real-world robotic imitation learning. Our framework utilizes a DINO object detection transformer to extract spatially-grounded object-query embeddings that serve as the conditioning signal for a diffusion-based policy. A primary contribution of this work is the systematic quantification of how scene complexity-measured via image entropy-affects robotic policy performance. By comparing rigid-object baselines with complex biological plant scenes, we demonstrate that organic morphology induces a measurable increase in pixel-level uncertainty that degrades standard pixel-centric models. Our results show that DINO + CDP mitigates this degradation by grounding action generation in stable object-level features. We evaluate our approach using a fully real-world manipulation dataset collected without simulation or synthetic pre-training. To isolate the impact of our architectural choices, we conduct a comparative study within a unified framework against convolutional (CNN-MLP), transformer-patch (ViT), and latent-variable (DETR + CVAE) variants. Experimental results in a robotic-arm biocell setup demonstrate that object-query-conditioned diffusion significantly improves task success rates, produces smoother trajectories, and exhibits superior robustness to high-entropy visual inputs, establishing a scalable pathway for imitation learning in challenging agricultural domains.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1829436"},"PeriodicalIF":3.7,"publicationDate":"2026-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13417565/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148621816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SpatialPrompting: pose-aware keyframe prompting for 3D spatial QA toward smart indoor environments.","authors":"Shun Taguchi, Hideki Deguchi, Takumi Hamazaki, Hiroyuki Sakai","doi":"10.3389/frobt.2026.1756174","DOIUrl":"10.3389/frobt.2026.1756174","url":null,"abstract":"<p><p>SpatialPrompting is a practical, training-free and model-agnostic framework for 3D spatial reasoning with off-the-shelf multimodal large language models, requiring no fine-tuning or 3D-specific inputs. It adopts a pose-aware, keyframe-driven prompting strategy: we select a compact, diverse set of frames using vision-language similarity, Mahalanobis distance, field of view, and image sharpness, and then verbalize each camera pose to enable multi-view, viewpoint-aware reasoning within a single structured prompt. Evaluations on ScanQA and SQA3D using GPT-4o and Gemini show that SpatialPrompting achieves competitive performance on ScanQA compared to existing approaches, while remaining slightly below the best fine-tuned methods on SQA3D under a training-free setting. On our Complex Spatial QA (CSQA) dataset, the proposed method improves accuracy from 62 to 78 (+16) over a GPT-4o baseline and consistently outperforms query-based keyframe selection. Furthermore, experiments across multiple models-including GPT-4o, Gemini, and Qwen-demonstrate that the framework generalizes across both proprietary and open-source models. By eliminating task-specific training and enabling reuse across different models without retraining, SpatialPrompting shifts the cost from model-specific training to flexible inference, offering a scalable and effective approach for real-world spatial reasoning as multimodal models continue to evolve.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1756174"},"PeriodicalIF":3.7,"publicationDate":"2026-07-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13407178/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148608692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Zero-shot semantic landmark-based visual odometry using foundation models for unstructured planetary exploration.","authors":"Cristina Pérez-Ramos, Leopoldo Altamirano-Robles, Miguel Chávez-Dagostino","doi":"10.3389/frobt.2026.1819714","DOIUrl":"10.3389/frobt.2026.1819714","url":null,"abstract":"<p><p>Precise autonomous navigation on unstructured planetary surfaces is a critical prerequisite for future exploration missions, particularly in GNSS-denied environments such as the Lunar South Pole or Martian deserts. Traditional Visual Odometry (VO) methods, which rely on tracking low-level geometric features (e.g., corners), often fail under the extreme illumination contrast of the Moon or the textural monotony of the Martian regolith. In this work, we present a zero-shot semantic landmark-based visual odometry approach that leverages the generalization capabilities of modern Foundation Models. Our approach uses the Segment Anything Model (SAM) to extract geological landmarks (rocks) and DINOv2 to generate view-invariant semantic descriptors that are matched across frames. We evaluate our pipeline across two distinct domains: a high-fidelity synthetic lunar environment (LuSNAR dataset) to test robustness against extreme lighting, and a real-world Martian analog dataset (Katwijk Beach) to assess sim-to-real transfer. Experimental results show that the proposed approach achieves a decimeter-level trajectory accuracy ( <math><mrow><mi>R</mi> <mi>M</mi> <mi>S</mi> <mi>E</mi> <mo>≈</mo></mrow> </math> 0.14 m) on the Martian analog and an <math><mrow><mi>R</mi> <mi>M</mi> <mi>S</mi> <mi>E</mi></mrow> </math> = 1.93 m on the most stable lunar traverse, without any domain-specific fine-tuning. Our results suggest that Foundation-Model-based semantic landmarks are a promising alternative to low-level features for zero-shot VO in planetary-like environments.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1819714"},"PeriodicalIF":3.7,"publicationDate":"2026-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13388163/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148563422","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"PALPABLE: enhanced bending stiffness and proprioception in soft actuators for laparoscopic palpation.","authors":"Christos Vasileios Lousis, Myrto Inglezou, Ilias Zournatzis, Georgios Violakis, Panagiotis Polygerinos","doi":"10.3389/frobt.2026.1832079","DOIUrl":"10.3389/frobt.2026.1832079","url":null,"abstract":"<p><p>The assessment of tissue properties via direct palpation is essential for localizing abnormalities during surgery. However, Minimally Invasive Surgery (MIS) eliminates this tactile feedback, creating a critical sensory gap. To address this, the EU Horizon project PALPABLE is developing novel fiber-optic sensing modalities for stiffness assessment. Successful integration of this technology requires an articulation interface that balances high maneuverability with the structural stability necessary for deflection-free tissue palpation. We present a single-degree-of-freedom (DOF) soft silicone actuator integrated with a modular, bioinspired passive stiffening element. Mechanical loading tests, motion tracking, and Finite Element Method (FEM)-guided design iterations validate the stiffening strategy and actuator performance. The design achieves a substantial increase in bending stiffness, improving the stiffness-pressure response by 8-fold relative to the unstiffened actuator under critical loading (normal force at a 90° bend). Furthermore, we developed a compact, custom pneumatic syringe pump with closed-loop pressure control to ensure precise, safe operation in surgical environments. By combining embedded fiber-optic sensing with machine-learning models, the system achieves free-space full-shape reconstruction with approximately 1% error. This work delivers a compact, load-capable, and sensorized soft actuator tailored for PALPABLE's probe manipulation requirements, enabling safe, controllable, and data-rich tissue palpation in confined anatomical spaces.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1832079"},"PeriodicalIF":3.7,"publicationDate":"2026-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13383039/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148608614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"OpenAI single-agent LLM architecture reduces computational overhead relative to multi-agent orchestration in a simulated mars rover decision-support benchmark.","authors":"Dan Sanabria","doi":"10.3389/frobt.2026.1877762","DOIUrl":"10.3389/frobt.2026.1877762","url":null,"abstract":"<p><p>Mars rover missions require decision-support systems that can interpret terrain, telemetry, environmental conditions, and mission objectives under delayed communication with Earth. This study evaluates whether multi-agent orchestration improves simulated Mars rover decision support compared with a single-agent baseline. A controlled benchmark of 100 synthetic mission-inspired rover scenarios was evaluated using OpenAI GPT-4o and GPT-5.5, with five repeated runs per scenario and architecture. Model-facing scenario inputs were separated from evaluator-side labels so that expected actions and hazards were reserved for scoring only. Performance was measured using decision accuracy, exact and substring-based semantic hazard F1, hazard error counts, latency, token usage, scenario-level paired statistical comparisons, and GPT-4o specialist-agent ablations. Across the tested OpenAI configurations, the single-agent architecture showed numerical advantages in decision accuracy and hazard-label alignment, but these decision-quality differences were not consistently significant under scenario-level statistical analysis with Holm-Bonferroni adjustment. The only decision-quality metric remaining significant was GPT-5.5 exact hazard F1, although absolute values were very low. The most reliable difference was computational efficiency: the single-agent architecture required substantially lower latency and token usage than the prompt-defined multi-agent orchestration architecture. Multi-agent orchestration generated broader hazard lists, including plausible non-canonical observations, but did not reliably improve aggregate decision accuracy or hazard F1. These findings suggest that, for short-context, tool-less, static decision-support tasks where all relevant context is available in a single input, multi-agent orchestration should be treated as a cost-bearing design choice rather than an assumed improvement. The study contributes a reproducible architecture-level benchmark for evaluating when LLM-based orchestration is worth its operational cost in mission-inspired workflows.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1877762"},"PeriodicalIF":3.7,"publicationDate":"2026-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13381255/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148608675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring deep reinforcement learning acceleration by superscaling data augmentation via branched fractal symmetries.","authors":"Ryan Vander Stelt, Cleiver Ruiz-Martinez, Caeden Rosen, Blake Hull, Juan Rojas","doi":"10.3389/frobt.2026.1791812","DOIUrl":"https://doi.org/10.3389/frobt.2026.1791812","url":null,"abstract":"<p><p>Learning deep reinforcement learning (DRL) policies directly in physical robots remains bottlenecked by slow wall-clock training times. We present preliminary research on <i>Branched Euclidean Group Fractal Symmetries</i>, a trajectory-level augmentation framework that super-scales group transformations to accelerate policy learning for manipulation. We model a Markov decision process (MDP) as a tree of state-action pairs; at each depth, affine transformations generate geometric structures, within which combinations of Euclidean group symmetries produce a large set of unique symmetric equivalences. A self-similar branching rule can also be iterated across depths to yield fractal symmetry expansions that populate the replay buffer with diverse yet consistent experiences with Cartesian or hand-centered image representations. We show the efficacy of our approach through simulated and real Franka robot manipulation contact tasks. Our study learns robust policies directly in physical robots in as fast as 14 min and in 20 min of wall-clock time on average. The code is accessible at https://github.com/lippyRobotics/fractalserl.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1791812"},"PeriodicalIF":3.7,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13369614/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148457572","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Morphology and control roles in perturbed standing recovery: a robotic study.","authors":"Yelin Jiang, Marc Murcia, Jitong Yang, Andre Seyfarth, Rolf Findeisen, Maziar A Sharbafi","doi":"10.3389/frobt.2026.1840011","DOIUrl":"10.3389/frobt.2026.1840011","url":null,"abstract":"<p><p>Postural balance is essential for both humans and robots, as failures increase fall risk and limit robotic performance in real-world settings. Although humans and robots share fundamental balancing mechanics, biological complexity limits the isolation of individual muscle functions and direct principle transfer to robots. In this study, we use EPA-Walker, a bio-inspired robot actuated by electric motors and pneumatic artificial muscles (PAMs), as a physical platform to investigate perturbed standing. Here, we focus on the PAM-driven actuation to systematically examine the roles of muscle morphology and control, and to validate biomechanical findings in a robotic setting. To enable a clear upper body perturbation, a Control Moment Gyroscope (CMG) was integrated. We evaluated two stabilization paradigms: passive standing, in which joint compliance was tuned through static PAM pressurization, and active balancing, in which a bio-inspired ground reaction force (GRF) feedback controller generated muscle reflexes. The results showed that biarticular thigh muscles, particularly the hamstrings and rectus femoris, played the most prominent role in enhancing robustness through both morphology in the passive experiments and reflex control in the active experiments, consistent with findings from human perturbation studies. While ankle muscles, such as the soleus, were essential for stable standing mainly through their passive morphological contribution, their reflex-based action could also improve robustness in a more specific manner through center-of-pressure regulation. Activating a single muscle could significantly improve robustness beyond morphology, enabling recovery from <math><mrow><mn>3</mn> <mtext> </mtext> <mi>N</mi> <mi>m</mi></mrow> </math> perturbations. Furthermore, synergistic reflex of biarticular muscles, especially hamstrings and gastrocnemius, extends the robustness to larger perturbations <math><mrow><mo>(</mo> <mrow><mn>5</mn> <mtext> </mtext> <mi>N</mi> <mi>m</mi></mrow> <mo>)</mo></mrow> </math> . Our contribution highlights the synchronization of control strategies with the underlying morphological design through a universal sensory feedback signal, namely, GRF. These findings demonstrate the value of bio-inspired robots as testbeds to understand the potential principles underlying human motor control and support the transfer of such principles to legged robots and assistive systems.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1840011"},"PeriodicalIF":3.7,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13368573/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148457600","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Vision-based vs. IMU-based upper-limb pose estimation in assisted dressing: a comparative study of positional accuracy and kinematic fidelity.","authors":"Yasmin Rafiq, Shenglin Wang, Mohammed Al-Nuaimi, Lyudmila Mihaylova, Robert M Hierons, Sanja Dogramadzi","doi":"10.3389/frobt.2026.1844439","DOIUrl":"10.3389/frobt.2026.1844439","url":null,"abstract":"<p><p>Accurate estimation of upper-limb kinematics is essential for applications such as rehabilitation assessment and assistive robotics, yet remains challenging in real-world scenarios involving occlusion and physical human interaction. While vision-based pose estimation methods have advanced significantly, their ability to recover reliable joint kinematics under such conditions remains unclear. This paper presents a systematic comparison of vision-based and wearable sensing approaches for upper-limb pose estimation during assisted dressing tasks. A monocular RGB-based convolutional neural network (CNN) and a temporally smoothed variant (CNN_temporal) are evaluated alongside a wearable IMU-based reconstruction method. All approaches are compared against an inverse kinematics (IK) reference derived from VICON motion capture data using participant-specific kinematic models. Performance is assessed using both positional error, measured via global and shoulder-centred mean per-joint position error (MPJPE), and kinematic agreement, measured via elbow flexion/extension angle error. Experiments on a real-world dataset of assisted dressing trials, involving an occupational therapist and three participants, demonstrate that IMU-based estimation provides consistently accurate and stable joint-angle reconstruction (e.g., <math><mrow><mo>∼</mo> <mn>12</mn> <mo>°</mo></mrow> </math> mean absolute error). In contrast, vision-based methods achieve reasonable positional accuracy (MPJPE <math><mrow><mo>∼</mo></mrow> </math> 0.20 m) but exhibit substantially larger errors in joint-angle estimation (often exceeding 80 <math><mrow><mo>°</mo></mrow> </math> ), particularly under occlusion. Temporal smoothing improves positional consistency but does not preserve kinematic fidelity. These results highlight a fundamental limitation of current vision-based approaches for tasks requiring accurate joint kinematics. The findings suggest that integrating inertial sensing or incorporating biomechanical constraints may be necessary to achieve reliable pose estimation in real-world assistive scenarios.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1844439"},"PeriodicalIF":3.7,"publicationDate":"2026-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13368569/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148457583","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}