Frontiers in Robotics and AI最新文献

{"title":"Adaptive dynamic programming for coordinated control of a dual-arm reconfigurable manipulator with unknown object constraints.","authors":"Zebin Ji, Yi Qin, Qiang Pan, Hucheng Jiang, Bing Ma","doi":"10.3389/frobt.2026.1923646","DOIUrl":"10.3389/frobt.2026.1923646","url":null,"abstract":"<p><p>Concentrating on various types of grasped objects, this work presents a distributed coordinated control scheme for a dual-arm reconfigurable manipulator (DRM) via adaptive dynamic programming (ADP). Kinematic and force analyses are performed to formulate the dynamics of a single-arm manipulator and the object using joint torque feedback technology and the Newton-Euler formulation. To address uncertainties associated with the grasped object, a gradient model-based adaptive algorithm is developed for online estimation of the grasp matrix relating the object to the DRM. By employing an adaptive observer to identify unknown dynamic terms, an enhanced optimal performance index is constructed that incorporates both object tracking performance and internal force effects within the manipulator. The performance index function is approximated by only a critic neural network, and the optimal control policy is obtained by policy iteration. Lyapunov stability analysis demonstrates that internal force error, position tracking error, and critic network weight approximation error are all uniformly ultimately bounded. Experimental results confirm the effectiveness of the developed coordinated control approach.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1923646"},"PeriodicalIF":3.7,"publicationDate":"2026-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13581869/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148992055","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":"Editorial: Robotic perception, multi-robot exploration, mapping, and autonomous computing.","authors":"Faiza Gul, Laith Abualigah, Rafal Szczepanski, Imran Mir","doi":"10.3389/frobt.2026.1974584","DOIUrl":"https://doi.org/10.3389/frobt.2026.1974584","url":null,"abstract":"","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1974584"},"PeriodicalIF":3.7,"publicationDate":"2026-09-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13581848/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148992029","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":"Editorial: Robotics in the performance, safety and learning of surgery - what next?","authors":"Sanju Lama, Hani J Marcus, Garnette R Sutherland","doi":"10.3389/frobt.2026.1927781","DOIUrl":"10.3389/frobt.2026.1927781","url":null,"abstract":"","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1927781"},"PeriodicalIF":3.7,"publicationDate":"2026-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13583154/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148991968","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":"Hierarchical path planning for low-altitude logistics UAVs in complex urban environments: an enhanced weighted A* approach with geometric smoothing.","authors":"Tan Lei, Shiming Liu, Yongtao Lei","doi":"10.3389/frobt.2026.1802470","DOIUrl":"10.3389/frobt.2026.1802470","url":null,"abstract":"<p><strong>Importance: </strong>Autonomous logistics UAVs operating in low-altitude urban airspace face a fundamentally distinct set of challenges from those in high-altitude flight-extreme obstacle density, regulatory geo-fencing, and unpredictable urban canyon wind shear demand path planners that are simultaneously fast, kinematically smooth, and safety-aware.</p><p><strong>Gap: </strong>Existing grid-based planners (e.g., standard A) generate geometrically suboptimal paths cluttered with sharp 45°/90° turns that waste energy and increase turning workload, while bio-inspired and sampling-based alternatives suffer from stochastic execution times and high hyperparameter sensitivity that preclude deterministic delivery scheduling. Critically, no prior deterministic planner simultaneously minimizes path length, cumulative turning angle, and obstacle proximity risk within a single unified cost function actually enforced during search. Methodology: We propose a hierarchical \"Search-and-Smooth\" framework comprising: (1) an Enhanced Weighted A algorithm whose edge-cost function directly encodes all three objectives-path length (Jlen), turning cost (Jturn), and spatial risk (<i>Jrisk</i>) - and whose inflated heuristic (<i>w = 1.5</i>) provides a provable <i>ε</i>-suboptimality bound while reducing node expansions; and (2) a greedy Line-of-Sight (LOS) post-processor that removes residual grid-discretization artifacts.</p><p><strong>Results: </strong>Monte Carlo simulations (<i>N = 50</i> randomized urban scenarios, obstacle density <i>ρobs = 20%</i>) demonstrate statistically significant improvements over standard A: computation time reduced by 42.8% ( <math><mrow><mi>p</mi> <mtext> </mtext> <mo><</mo> <mtext> </mtext> <mn>0.001</mn></mrow> </math> , Wilcoxon signed-rank <math><mrow><mi>W</mi> <mo>=</mo> <mn>1275</mn></mrow> </math> ), cumulative turning angle reduced by 69.9% ( <math><mrow><mi>p</mi> <mtext> </mtext> <mo><</mo> <mtext> </mtext> <mn>0.001</mn></mrow> </math> ; <math><mrow><mi>W</mi> <mo>=</mo> <mn>1275</mn></mrow> </math> ), and path length maintained within 1.8% of the standard A baseline despite the sub-optimal heuristic inflation. Comparisons against Theta* and a representative PSO baseline confirm the competitive advantages of the proposed method for deterministic, low-latency logistics deployment within the tested synthetic simulation conditions.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1802470"},"PeriodicalIF":3.7,"publicationDate":"2026-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13563452/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148956516","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":"Robust trajectory tracking control of an underwater robotic manipulator using an IPSO-PID framework under hydrodynamic and ocean-current disturbances.","authors":"Jinming Yao, Jiahang Yang, Chunxu Li","doi":"10.3389/frobt.2026.1915434","DOIUrl":"https://doi.org/10.3389/frobt.2026.1915434","url":null,"abstract":"<p><p>Robust trajectory tracking is a fundamental requirement for underwater robotic manipulators operating in disturbed marine environments. However, hydrodynamic damping, added mass, buoyancy effects, multi-joint coupling, and external ocean-current disturbances can significantly degrade joint tracking accuracy and control stability. To address this problem, this paper proposes an improved particle swarm optimization-based PID (IPSO-PID) control framework for robust trajectory tracking of a 6-DOF underwater manipulator. A UR5-type underwater manipulator is modeled using the D-H parameter method, while its dynamic behavior is formulated based on the Lagrange method. To represent underwater environmental effects, the Morison equation is introduced to describe fluid drag and added-mass forces, and a compound ocean-current disturbance model consisting of a constant bias load and a low-frequency periodic component is further constructed. In the control framework, the standard PSO algorithm is improved by incorporating a linearly decreasing inertia weight and adaptive learning factors, thereby enhancing global exploration in the early optimization stage and local convergence in the later stage. A comprehensive fitness function is designed by jointly considering trajectory tracking error, error convergence behavior, and control effort, enabling coordinated optimization of the PID parameters for all six joints. Comparative simulations are conducted in MATLAB/Simulink using conventional PID, fuzzy PID, standard PSO-PID, and the proposed IPSO-PID controller. The results show that the proposed controller achieves better tracking accuracy and stronger disturbance rejection capability than the comparison methods. Under compound ocean-current disturbance, the overall mean absolute error across all six joints is reduced from approximately 0.0044 rad with the conventional PID controller to approximately 0.0014 rad with the IPSO-PID controller, corresponding to a reduction of about 68.2%. These results indicate that the proposed control framework can effectively improve the stability and robustness of underwater manipulator trajectory tracking, providing a feasible solution for underwater robotic tasks such as inspection, grasping, maintenance, and marine intervention.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1915434"},"PeriodicalIF":3.7,"publicationDate":"2026-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13561991/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148956582","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":"Ambiguous definitions of industrial mobile robots: a survey of German-speaking experts.","authors":"Sven Franke, Jan-Bernd Igelmann, Jérôme Rutinowski, Markus Pauly","doi":"10.3389/frobt.2026.1837981","DOIUrl":"10.3389/frobt.2026.1837981","url":null,"abstract":"<p><p>Industrial mobile robots are increasingly deployed in warehouses and production sites and interaction takes place with different stakeholders from different disciplines. In such sociotechnical settings, terminology shapes mental models and perceptions that influence trust, interaction, and safe collaboration. Key labels such as driverless transport vehicle (DTV), automated guided vehicle (AGV), autonomous mobile robot (AMR) and mobile robot (MR) are used inconsistently across standards, industry practice and academic literature. It remains unclear how experts actually interpret these terms in professional communication, since the terms are only in part defined by standards and perceived differently throughout the industry. We address this gap with an online survey of 53 mobile robotics experts. Spontaneous associations, the perceived distinctiveness between pairs of terms, consistency with general terminology statements and attributed technical characteristics are identified. The results are contrasted with normative references and interoperability initiatives. The findings indicate systematic, term-specific perceptions. Participants report divergent perceptions, conflicting definitions and a lack of uniform terminology. Experts rated AGV and DTV as largely indistinguishable (relative effect = -0.51/-1), but considered AGV and AMR clearly more separable (0.37/1), while MR functioned as a broad umbrella term that evoked capability-related associations yet remained ambiguous. Based on these ambiguities, we synthesize recommendations for how terminology should be communicated and qualified in industrial mobile robotics to reduce expectation gaps across stakeholder groups and facilitate a joint, uniform understanding.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1837981"},"PeriodicalIF":3.7,"publicationDate":"2026-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13561907/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148956543","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":"An online survey of encouragement strategies for older adults' daily worries: a conversational robot versus a human partner.","authors":"Lingxuan Xiang, Marin Nishimura, Hideaki Kikuchi","doi":"10.3389/frobt.2026.1843519","DOIUrl":"10.3389/frobt.2026.1843519","url":null,"abstract":"<p><p>Conversational robots are increasingly proposed to provide everyday emotional support for older adults. However, it remains unclear which encouragement strategies are perceived as effective and whether interpersonal strategy preferences transfer to Human-Robot Interaction (HRI). This study compared the perceived effectiveness of five encouragement strategies when delivered by a conversational robot versus a human partner, examined moderation by worry type, and explored associations with recipients' emotional states. Japanese young-old adults aged 65-74 (final N = 55) completed an online scenario-based survey. Participants were randomly assigned to a human or conversational robot condition, each combining an identity label with distinct voice characteristics (a human-recorded voice or a robot-recorded voice). They recalled one physical and one psychological daily worry and rated perceived encouragement for five strategy types: reassurance and affirmation, expressing concern, encouragement to act, offering specific actions, and distraction. Within-subject standardized ratings were analyzed using linear mixed-effects models with a participant random intercept. Positive and Negative Affect Schedule (PANAS) were examined exploratorily. Results showed that perceived effectiveness differed by strategy type and notably by condition. In the human condition, offering specific actions showed the highest relative effectiveness, with descriptively clearer differentiation among strategies. In the robot condition, reassurance and affirmation ranked highest, and differentiation among strategies was descriptively reduced. Distraction was consistently least preferred. The present study did not observe a statistically significant effect of worry type on the overall ranking patterns. Exploratory analyses suggested that positive affect was associated with encouragement evaluation at the raw-score level in the robot condition, whereas associations involving negative affect did not survive correction for multiple comparisons and are therefore interpreted descriptively rather than as statistically significant findings. These findings suggest that interpersonal encouragement strategy preferences may not directly transfer to human-robot interaction as operationalized in the present study. The observed differences between conditions may reflect the combined effect of the identity label and voice characteristics rather than the robot's identity alone. The results highlight the need for partner-specific, context-adaptive encouragement design in conversational robots for older adults.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1843519"},"PeriodicalIF":3.7,"publicationDate":"2026-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13554955/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148921432","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":"Encoder-free hexapod dead reckoning on sand using learned inertial correction.","authors":"Mingyu Pan, Maxwell A Rollins, Samuel H Yang, Kathryn A Daltorio","doi":"10.3389/frobt.2026.1920986","DOIUrl":"10.3389/frobt.2026.1920986","url":null,"abstract":"<p><p>For inexpensive legged robots traversing remote natural terrain, minimal-sensing localization strongly affects autonomous range and search efficiency. Although proprioceptive sensors such as inertial measurement units (IMU) are compact and widely used on robotic platforms, direct integration of noisy inertial measurements causes small sensor errors to propagate into large pose estimation errors, limiting their use as independent dead reckoning sensors. In this work, we address drift in encoder-free dead reckoning by combining body-frame acceleration profiles from a compact multi-IMU array with commanded motion parameters in a convolutional neural network (CNN) bias estimator. The estimator predicts the difference between commanded and realized platform displacement, without using joint encoders or explicit terrain labels. In outdoor local search experiments with an encoder-free hexapod robot, the proposed approach reduces mean dead reckoning drift by approximately 41% on grass and 59% on sand relative to an idealized commanded-kinematics baseline. These results indicate that a platform-specific dataset collected in controlled indoor terrains improves outdoor dead reckoning performance, even when the deployment terrain is not fully represented in the training dataset. This method provides a low-cost, compact proprioceptive drift reduction method for minimally instrumented legged robots when exteroceptive localization is unavailable due to cost or environmental conditions. We expect this approach to be helpful in developing future low-cost robots for scrambling across extraterrestrial, underwater, or underground granular media where leg slip is both common and hard to predict.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1920986"},"PeriodicalIF":3.7,"publicationDate":"2026-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13553437/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148921386","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":"Bayesian selective scan for robust long-horizon action recognition via reliability-aware discretization.","authors":"Chunjie Wang","doi":"10.3389/frobt.2026.1786524","DOIUrl":"https://doi.org/10.3389/frobt.2026.1786524","url":null,"abstract":"<p><strong>Introduction: </strong>Long-horizon action recognition is increasingly deployed in streaming and safety-critical settings, where short but severe corruptions, such as motion blur or adversarial perturbations, can overwrite temporal memory and cause errors to persist after unreliable segments end. Selective State Space Models (SSMs) enable linear-time inference over long contexts, but deterministic discretization can amplify this memory-contamination failure mode.</p><p><strong>Methods: </strong>We propose BayesMamba, a reliability-aware selective SSM that integrates Bayesian uncertainty into SSM discretization. BayesMamba uses Bayesian Selective Scan (B-Scan) to estimate per-frame aleatoric and epistemic uncertainty, convert uncertainty into a reliability signal, and modulate both the effective time step and input injection so that state updates are suppressed when observations are unreliable. We evaluate BayesMamba on Kinetics-400 (K400) and Something-Something V2 (SSv2).</p><p><strong>Results: </strong>Under clean evaluations, BayesMamba achieves comparable or slightly improved performance, with 82.4% top-1 accuracy on K400 and 68.7% on SSv2, compared with 82.1% and 68.2% for a deterministic selective SSM baseline, while retaining similar computational cost. On SSv2 under bursty motion blur, BayesMamba improves robust accuracy from 51.4% to 54.2%; under projected gradient descent (PGD, ε = 4/255, ε = 8/255, and ε = 16/255), it improves robust accuracy from 18.2% to 19.8%. BayesMamba also reduces post-corruption persistence, decreasing Persist(64) from 0.41 to 0.19 and Persist(128) from 0.38 to 0.16. Under distribution shift, it improves calibration, reducing corrupted-set ECE from 0.158 to 0.097 and corrupted-set NLL from 2.34 to 1.66.</p><p><strong>Discussion: </strong>These results indicate that reliability-aware discretization mitigates memory contamination and improves long-horizon robustness with modest overhead while preserving clean action recognition accuracy.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1786524"},"PeriodicalIF":3.7,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13547921/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148907671","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":"Distributed environmental sensing and soil moisture gradient navigation in modular robots.","authors":"Henry Alberto Hernandez Martinez, Holman Montiel Ariza, Luis Fernando Pedraza Martinez","doi":"10.3389/frobt.2026.1895019","DOIUrl":"10.3389/frobt.2026.1895019","url":null,"abstract":"<p><p>Agricultural environments exhibit spatial variations in soil moisture that provide directional information for exploration and monitoring tasks. This work presents a distributed navigation strategy for modular robots based on local soil moisture gradients and neighbor-to-neighbor information exchange. The proposed architecture separates navigation and locomotion into two coordinated layers. The navigation layer estimates movement direction from environmental measurements and obstacle perception. The locomotion layer transforms directional decisions into coordinated motion through coupled oscillators. The proposed method was implemented on an EMeRGe modular robotic platform and was evaluated through 360 physical experiments conducted with three modular configurations under heterogeneous soil moisture conditions. The experimental evaluation compared the proposed method with Hill Climbing, Ant Colony Optimization, and Genetic Algorithm navigation strategies operating under identical sensing and locomotion conditions. The obtained results showed reductions in convergence time, decreases in directional estimation error, increases in navigation efficiency, and lower velocity variance across the evaluated scenarios. These results indicate that local soil moisture gradients can be used to estimate navigation directions without global maps or centralized control. This approach establishes a relationship between environmental sensing, distributed directional estimation, and motion generation in modular robotic systems for agricultural monitoring applications.</p>","PeriodicalId":47597,"journal":{"name":"Frontiers in Robotics and AI","volume":"13 ","pages":"1895019"},"PeriodicalIF":3.7,"publicationDate":"2026-08-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13587347/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149009244","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}