Robotics and Computer-integrated Manufacturing最新文献

{"title":"Grid-to-Robot: Deep Wasserstein generative modeling of robot/power grid interaction using hybrid adversarial Residual Networks","authors":"Ashkan Safari ,&nbsp;Hamed Kharrati ,&nbsp;Afshin Rahimi ,&nbsp;M. Ali Tavallaei","doi":"10.1016/j.rcim.2025.103086","DOIUrl":"10.1016/j.rcim.2025.103086","url":null,"abstract":"<div><div>Smart Manufacturing (SM) is an important factor for driving innovation, enhancing operational efficiency, and increasing sustainable industrial growth in an increasingly competitive and resource-constrained world. However, it faces several challenges related to increasing energy consumption and climate change. The high energy demands of connected devices and robotic manipulators increase the carbon footprint. To resolve this issue, most enterprises are now transitioning to use Renewable Energy Sources (RES), and optimizing their power and energy usage, while holding the process efficient. To fully achieve this transition, a detailed power modeling of the robotic manufacturing system is crucial and, therefore, it is important to investigate this power modeling of the robotic manipulators’ consumption in a Smart Sustainable Manufacturing (SSM) to achieve the best power modeling results and better integrability analytics in optimal power planning of the robotic systems power supply. To this end, this paper presents a deep Generative Artificial Intelligence (GAI)-based modeling of robotic manipulators’ power supply interaction with the power grid, and RES. In the proposed system, which is powered by solar energy and the power grid, a SSM equipped with ten 6-Degrees of Freedom (DoF) robotic manipulators is considered in the presence of Battery Energy Storage Systems (BESSs). Subsequently, a Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP) is employed to generate synthetic data for the system alongside the real data, thereby expanding the analytical horizons across varying operational characteristics of the system. Following this, a Residual Networks (ResNet) is developed to comprehensively analyze and predictively model the power consumption of the manipulators and their interactions with the power supply resources. Finally, the proposed hybrid GAI modeling strategy is numerically evaluated across a broad spectrum of Key Performance Indicators (KPIs) (MSE= <span><math><mrow><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>4</mn></mrow></msup></mrow></math></span>, MAE= <span><math><mrow><mn>3</mn><mo>.</mo><mn>6</mn><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>3</mn></mrow></msup></mrow></math></span>, R²= 99.98%, MARE= <span><math><mrow><mn>1</mn><mo>.</mo><mn>97</mn><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>2</mn></mrow></msup></mrow></math></span>, RMSPE= <span><math><mrow><mn>8</mn><mo>.</mo><mn>83</mn><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>2</mn></mrow></msup><mtext>%</mtext></mrow></math></span>, MSRE= <span><math><mrow><mn>7</mn><mo>.</mo><mn>8</mn><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>3</mn></mrow></msup></mrow></math></span>, RMSRE= <span><math><mrow><mn>8</mn><mo>.</mo><mn>84</mn><mo>×</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>2</mn></mrow></msup></mrow><","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103086"},"PeriodicalIF":9.1,"publicationDate":"2025-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144596500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing industrial human action recognition framework integrating skeleton data acquisition, data repair and optimized graph convolutional networks","authors":"Bojian Liu ,&nbsp;Yufeng Yao,&nbsp;Honggang Wang ,&nbsp;Zengmin He,&nbsp;Anyang Dong","doi":"10.1016/j.rcim.2025.103089","DOIUrl":"10.1016/j.rcim.2025.103089","url":null,"abstract":"<div><div>The precise interpretation of human actions is crucial for seamless interaction and operational efficiency for industrial human-robot collaboration. However, existing skeleton-based action recognition methods focus on algorithmic applications while overlooking key challenges such as robust data acquisition, validation, and repair. Additionally, the scarcity of high-quality industrial datasets and the challenges in distinguishing similar actions further limit the capability to infer operators' intentions accurately. This paper presents a novel framework to address challenges utilizing integrating skeleton data acquisition, effective data augmentation method, and an optimized graph convolutional network. Specifically, the proposed framework employs a pose estimation method for 2D (two-dimensional) joint estimation and a 2D-to-3D (three-dimensional) lifting technique, supplemented with a robust method for repairing invalid skeleton data and a skeletal feature-based data augmentation strategy. To enhance action recognition, this paper introduces the Channel-Topology Refinement Graph Convolutional Network Plus (CTR-GCN-Plus), which incorporates dynamic topology learning and multi-channel feature aggregation, augmented with hand motion integration for finer differentiation of similar actions. The proposed framework is evaluated on an industrial assembly dataset incorporating challenging scenarios, such as occlusions and similar actions. Experimental results demonstrate that the proposed methods significantly improve accuracy, enhance recognition of similar actions, and effectively account for individual variations, outperforming existing approaches in industrial human-robot collaboration environments.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103089"},"PeriodicalIF":9.1,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144587573","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A modeling and adaptive evolution method for simulation parameters of digital twin shop floor","authors":"Litong Zhang ,&nbsp;Yu Guo ,&nbsp;Shengbo Wang ,&nbsp;Guanguan Zheng ,&nbsp;Weiwei Qian ,&nbsp;Shaohua Huang ,&nbsp;Weiguang Fang","doi":"10.1016/j.rcim.2025.103090","DOIUrl":"10.1016/j.rcim.2025.103090","url":null,"abstract":"<div><div>Digital twin (DT) model can accurately predict the future state of the shop floor and promptly identify potential problems, abnormal situations, or optimization opportunities. However, traditional production simulation method without considering the temporal characteristics of entities’ attributes. In the life cycle of physical entities, its attributes’ change will increase the DT simulation parameters’ error. Therefore, deep learning algorithms are used to model and evolve the simulation parameters of the digital twin shop floor (DTSF) to improve simulation accuracy. Firstly, the interaction mechanism between deep learning and discrete event simulation is designed. Then, a sequential regression variational autoencoder (SRVAE) is proposed to model the DT temporal parameters. Furthermore, the online instructor algorithm is proposed to update SRVAE through online data. This approach improves the simulation accuracy of DTSF while allowing its parameters to be self-maintained. And the effectiveness of the proposed method is verified by a case study.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103090"},"PeriodicalIF":9.1,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144523318","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Robotic abrasive cloth flap wheel polishing system and multivariate parameter decision-making method for blade leading and trailing edges","authors":"Dongbo Wu ,&nbsp;Huilin Li ,&nbsp;Hui Wang ,&nbsp;Suet To","doi":"10.1016/j.rcim.2025.103075","DOIUrl":"10.1016/j.rcim.2025.103075","url":null,"abstract":"<div><div>The surface quality of the blade leading and trailing edges (LTE) impacts jet-engine performance. This study constructs a robotic abrasive cloth flap wheel (ACFW) polishing system and proposes a multivariate parameter decision-making method for the optimal polishing surface quality of the blade LTE. The robotic polishing system, including machine vision, offline programming, and constant force control, is first developed, and the blade polishing process, including blade clamping, on-machine measurement, position compensation, and polishing strategies, is then analyzed. Finally, a multivariate parameter decision-making method is proposed based on the surface roughness regression model (RM) and adaptive genetic algorithm-backpropagation (AGA-BP) network. The surface roughness RM, influencing factors, and the response curve are determined through a full factorial design (FFD) and the response surface methodology (RSM). Meanwhile, the AGA-BP network, which integrates the adaptive genetic algorithm (AGA) and backpropagation neural network (BPNN), is proposed to model and predict the roughness of the blade surface. Based on the optimal parameters, the surface roughness of the blade LTE will reach Ra = 0.142μm, which illustrates that the developed robotic polishing system is highly efficient and feasible. Furthermore, the mean error percentages of the RM, BPNN, and AGA-BP predictions are 17.946%, 9.633%, and 1.495%, respectively, for four random test datasets. The maximum error for the AGA-BP network is 1.995%, while the minimum is 0.758%. This network model can accurately predict surface roughness for the robotic polishing system of the Ti-6Al-4V blade LTE.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103075"},"PeriodicalIF":9.1,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144523317","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An efficient trochoidal toolpath for freeform surface polishing with uniform material removal and isotropic texture","authors":"Hai-Long Xie ,&nbsp;Qing-Hui Wang ,&nbsp;Yue-Feng Li ,&nbsp;Jing-Rong Li ,&nbsp;A.Y.C. Nee ,&nbsp;S.K. Ong","doi":"10.1016/j.rcim.2025.103087","DOIUrl":"10.1016/j.rcim.2025.103087","url":null,"abstract":"<div><div>Isotropic polished surface topography (PST) and uniform material removal (MR) are the keys to achieving good polished surface quality and favorable surface functional properties. However, few methods reported thus far can ensure obtaining uniform MR, isotropic PST, and high polishing efficiency simultaneously for freeform surfaces. Therefore, this paper presents a novel trochoidal toolpath enabling simultaneous optimization of polishing efficiency, PST isotropy, and MR uniformity for freeform surfaces. With this approach, a new polishing toolpath model is proposed by combining an adaptive trochoid and an improved Hilbert curve, which has more adjustable parameters, better multi-directionality, higher randomness, and smoothness. Then, a polishing performance index, namely “PSD (Power Spectral Density) roundness”, is proposed to quantitatively characterize the PST isotropy. Based on this, a toolpath optimization algorithm with the comprehensive consideration of polishing efficiency, MR uniformity, and PST isotropy is further proposed. Finally, the proposed method was compared with existing toolpath planning methods through simulation and physic polishing experiments to demonstrate its effectiveness and superiority. The result shows that the polishing efficiency, MR uniformity, and PST isotropy are comprehensively improved by nearly 30 %, 49 %, and 20 times respectively when compared to the existing methods, which is of great significance in improving the polished surface quality for freeform surfaces.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103087"},"PeriodicalIF":9.1,"publicationDate":"2025-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144502234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Posture optimization for improving the kinematics performance of robotic polishing under combined constraints by using a KC-ADP method","authors":"Yingpeng Wang,&nbsp;Huipeng Song,&nbsp;Haoyu Fu,&nbsp;Yuwen Sun","doi":"10.1016/j.rcim.2025.103084","DOIUrl":"10.1016/j.rcim.2025.103084","url":null,"abstract":"<div><div>The flexibility of industrial robots in posture adjustment has driven their widespread adoption in polishing applications, providing an effective means to enhance machining performance. To satisfy the growing industrial demand for high-quality and high-efficiency machining of complex surfaces, posture optimization must address multiple constraints, including interference-free operation, singularity avoidance, stiffness performance index limits, and the kinematic parameter limits. The simultaneous consideration of all these factors poses a significant challenge, and existing methods do not adequately address the global optimality and the convergence of the solution process. This paper proposes a novel posture optimization model to improve the comprehensive performance of robotic postures by integrating enhancements in both kinematic and stiffness performance under fundamental geometric constraints. An efficient and stable algorithm, designated as Kinematics-Constrained Adaptive Dynamic Programming (KC-ADP), is developed to solve the optimization problem. First, the combined constraints are modeled based on the robotic polishing system. Next, feasible posture solutions corresponding to different redundant parameters are collected according to machining requirements and transformed into a directed graph using the proposed Multi-Constraints Search Space Generation (MCSSG) algorithm. The optimal posture sequence is then obtained through adaptive dynamic programming, ensuring the availability of feasible postures at each step and resolving the conflict between multi-order kinematic constraints and the objective function. A series of simulations and experiments were conducted to validate the proposed method and the results demonstrate that the proposed approach significantly improves machining performance.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103084"},"PeriodicalIF":9.1,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144490009","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Robotic jamming-free and precise assembly for peg-in-hole operation by multi-DOF force-controlled parallel end-effector","authors":"Qixiang Zhao ,&nbsp;Hailemichael Yilma Hailegebrial ,&nbsp;Wei Wang ,&nbsp;Ke Wen ,&nbsp;Lianyu Zheng","doi":"10.1016/j.rcim.2025.103083","DOIUrl":"10.1016/j.rcim.2025.103083","url":null,"abstract":"<div><div>In smart manufacturing, the precision peg-in-hole assembly under contact-unknown conditions poses a major challenge, which has spurred growing interest in enhancing active compliance in robotic assembly systems. Multi-Degree-of-Freedom (Multi-DOF) compliance provides high adaptability for addressing this assembly challenge. However, problems such as excessive size, insufficient dynamics, and complex jamming mechanisms have emerged. To solve these problems, we propose a jamming-free strategy. This strategy uses a lightweight, compact multi-DOF end-effector with a 3-Prismatic-Revolute-Spherical (3-PRS) parallel configuration, along with an active force control algorithm. The force control result of an absolute mean error of 1 N is observed for a setpoint value of 100 N along the z-axis. Meanwhile, the moments about the x- and y-axis are controlled within an absolute mean error of 0.04 N.mm. With the accurate compliance provided by the proposed force-controlled end-effector, the peg-in-hole operation is considered as a reciprocal transition between a transient one-point contact and a steady two-point jamming. Based on admittance control, a control strategy is designed to transform the two-point jamming into one-point contact, thereby preventing jamming during the precision peg-in-hole operation with significant uncertainties, eliminating the need for repeated positional adjustments during insertion. Experimental results show that the peg can be inserted into different holes with various pose errors and unknown dimensional tolerances. These results validate the rationality of the proposed jamming-free strategy for precision peg-in-hole assembly. The jamming-free peg-in-hole assembly employing a force-controlled parallel end-effector exhibits superior robustness and operational stability when it is applied into a satellite assembling task.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103083"},"PeriodicalIF":9.1,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144471502","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Integrated dynamic scheduling method for hybrid flow shop with machine preventive maintenance based on cooperative multi-agent deep reinforcement learning","authors":"Siqi Liu ,&nbsp;Haiping Zhu ,&nbsp;LieZheng Sheng","doi":"10.1016/j.rcim.2025.103085","DOIUrl":"10.1016/j.rcim.2025.103085","url":null,"abstract":"<div><div>Hybrid flow shop widely used in manufacturing industry is facing the challenge of complex and dynamic production environment. Current study mostly cannot consider the machine preventive maintenance and dynamic events in hybrid flow shop scheduling process. Therefore, this paper presents an integrated dynamic scheduling method for hybrid flow shop scheduling problem- unrelated parallel machine considering preventive maintenance (DHSFP-UPM-PM). And the multi-scheduling objectives include minimizing completion time, processing cost and maintenance cost. Firstly, the definition of research problem and the basic maintenance strategy are presented in detail. And an integrated mathematic model of DHSFP-UPM-PM is constructed. Then the integrated dynamic scheduling framework based on cooperative multi-agent deep reinforcement learning is proposed to solve the DHSFP-UPM-PM. Based on the above, we proposed a cooperative multi-processing stage agents (PSA) approach to realize the transformation from traditional single-agent to multi-agent. Meanwhile, the cooperative multi-agent Markova Decision Process is formulated to clarify the interaction between each agent and production environment. The state and action space as the key elements of scheduling model is also designed for each PSA. To optimize scheduling objectives, this paper further formulates new global reward mechanism and centralized training-decentralized execution method based on multi agent proximal policy optimization. Lastly, the experiment results verify the superiority and effectiveness of the proposed method when solving integrated scheduling problem and dynamic event. And the proposed method presents remarkable adaptability and flexibility under a different production scenario which prove the benefits of multi-agent deep reinforcement learning in complex and dynamic environment.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103085"},"PeriodicalIF":9.1,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144472090","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Autogeneration and optimization of pick-and-place trajectories in robotic systems: A data-driven approach","authors":"Mingkun Wu ,&nbsp;Alisa Rupenyan ,&nbsp;Burkhard Corves","doi":"10.1016/j.rcim.2025.103080","DOIUrl":"10.1016/j.rcim.2025.103080","url":null,"abstract":"<div><div>For manufacturing processes in industries such as aerospace, automotive and electronics, it is essential for robots to perform pick-and-place tasks with high efficiency and accuracy. To this end, we propose a data-driven framework to generate a pick-and-place trajectory that ensures high-accuracy tracking while simultaneously reducing residual vibration, which is particularly valuable for commercial industrial robots with unchangeable control systems. The proposed approach includes both the trajectory generation and the trajectory compensation phases. In the first phase, we plan a pick-and-place trajectory that effectively attenuate residual vibration by minimizing the acceleration energy within a specific frequency spectrum, where the frequency parameters and the time ratio are tuned by Bayesian optimization. In the second phase, we focus on improving the tracking accuracy by incorporating a trajectory compensation term. More precisely, we first learn a Koopman operator-based linear predictor, where a model-agnostic meta-learning framework is introduced to mitigate the demand for massive data from the target system. Then, we calculate the trajectory compensation term using an iterative learning control-based method. The proposed methodology is entirely data driven, enabling its application in various robotic systems and has potential in other manufacturing applications. We demonstrate the approach through high-fidelity simulations on Delta robots – a representative parallel robot, where trajectory generation effectively removes vibrations, and through physical experiments on UR5 robots – a typical serial robot. The results of the experiment show that the positioning accuracy of the three joints of the UR5 robot improved by 94%, 43%, and 96%.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103080"},"PeriodicalIF":9.1,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144365296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An adaptive disturbance compensation method for force-sensorless control systems applied to robotic milling","authors":"Han-Hao Tsai,&nbsp;Jen-Yuan Chang","doi":"10.1016/j.rcim.2025.103082","DOIUrl":"10.1016/j.rcim.2025.103082","url":null,"abstract":"<div><div>This paper introduces a novel disturbance compensation method for a robust, active-force-controlled end-effector. This system integrates a disturbance observer (DOB) and a reaction force observer (RFOB) while employing a numerical optimization-based extremum-seeking algorithm. Conventional force/torque sensors, when employed in serially connected manipulator systems, often trigger unstable responses due to the presence of non-collocated modes. Furthermore, conventional reaction force observers may fail to accurately estimate the contact force between the robot and its environment when external disturbance terms are not perfectly modeled a priori. In response to these challenges, recent research has reintegrated force/torque sensors into the control architecture, employing filter-based sensor fusion techniques to mitigate disturbance effects. However, these approaches fail to address the inherent stability challenges caused by the mounting and serial connection of force/torque sensors within the robot manipulator system, which in turn increases the design complexity of reaction-force-observer-based force control systems. To overcome these limitations, this paper proposes adaptive disturbance compensation methods that leverage position feedback information related to the depth of cut in milling processes. The proposed method adaptively compensates for time-varying disturbances, such as tool wear and abrupt feed rate changes, ensuring consistent performance under dynamic conditions. Compared to conventional position-controlled and force-controlled methods, the proposed approach exhibits improved robustness, precision, and versatility, positioning it as a promising solution for advancing robotic milling technologies toward practical industrial applications.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"97 ","pages":"Article 103082"},"PeriodicalIF":9.1,"publicationDate":"2025-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144313478","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}