Robotics and Computer-integrated Manufacturing最新文献

{"title":"Real-time response to machine failures in self-organizing production execution using multi-agent reinforcement learning with effective samples","authors":"Yong Gui ,&nbsp;Dunbing Tang ,&nbsp;Yuqian Lu ,&nbsp;Haihua Zhu ,&nbsp;Zequn Zhang ,&nbsp;Changchun Liu","doi":"10.1016/j.rcim.2025.103038","DOIUrl":"10.1016/j.rcim.2025.103038","url":null,"abstract":"<div><div>With the growing demand for personalized production, multi-agent technology has been introduced to facilitate rapid self-organizing production execution. The application of communication protocols and dynamic scheduling algorithms supports multi-agent negotiation and real-time scheduling decisions in response to conventional production events. To address machine failures, real-time response strategies have been developed to manage jobs affected by the disruptions. However, the performance of existing strategies varies significantly depending on the real-time production state. In this paper, we propose a real-time response strategy using multi-agent reinforcement learning (MARL) that provides an appropriate response strategy for each job affected by machine failures, considering the real-time production state. Specifically, we establish a self-organizing production execution process with machine failures to specify the real-time response problem. Subsequently, a Markov game involving multiple buffer agents is constructed, transforming the real-time response problem into a MARL task. Furthermore, a continuous variable ranging from 0 to 1 is defined as the action space for each buffer agent, allowing it to select a response strategy for each affected job. Finally, a modified multi-agent deep deterministic policy gradient (MADDPG) algorithm is introduced, leveraging effective samples to train buffer agents at each failure moment. This enables the selection of an optimal response strategy for each affected job. Experimental results indicate that the proposed real-time response strategy outperforms both existing response strategies and the original MADDPG-based strategy across 54 distinct production configurations.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103038"},"PeriodicalIF":9.1,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868269","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":"Extremely constrained robotic milling posture active adjustment on curved surface: Divide and conquer strategy with index of generalized accumulative work","authors":"Shengqiang Zhao ,&nbsp;Fangyu Peng ,&nbsp;Juntong Su ,&nbsp;Xiaowei Tang ,&nbsp;Teng Zhang ,&nbsp;Jiawei Wu ,&nbsp;Rong Yan","doi":"10.1016/j.rcim.2025.103036","DOIUrl":"10.1016/j.rcim.2025.103036","url":null,"abstract":"<div><div>Industrial robots have been widely used in many manufacturing scenarios, including curved surface with complicated geometric features, due to their flexible capability of posture adaptation and large machining range. This work utilizes the multi-solution feasible space characteristic of robot to construct a novel milling posture active adjustment model under multi-source heterogeneous constraints. Firstly, a generalized potential field is developed for dimensional unification, after analyzing the multi-source heterogeneous constraints from the robot ontology (macro-view) to the cutter-workpiece engagement (micro-view). In terms of the extremely constrained process requirements on the curved surface, a framework to solve the global posture planning in robotic milling process is proposed. For free-form surfaces with uneven curvature distribution, surface division is developed based on curvature similarity, under the premise of high-order continuity on subdivision boundary. A latent dynamic response model is established for active posture adjustment by introducing the imaginary potential torque. Significantly, generalized accumulative work is innovatively proposed as a global index for the intelligent optimization of milling posture in multi-solution feasible space. In comparison with the commercial planning result and the proposed active posture adjustment strategy, milling experiments are carried out on the complex parametric surface. Under the evaluation both in the distribution characterization (cell perspective) and eigenvalue quantization (string perspective), the milling results of the proposed strategy has improvement on four kinematic performance, and prominent breakthrough in machining precision by 12.12 % improvement. The active posture adjustment strategy for global optimization proposed in this work shows great application potential in the extremely constrained scenarios of robotic milling field, especially restricted working space, large machining range and high milling precision requirements.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103036"},"PeriodicalIF":9.1,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143854594","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":"Optimization of machine configuration and scheduling in the hybrid flow shop using a linear programming-driven evolutionary approach","authors":"Mengya Zhang ,&nbsp;Cuiyu Wang ,&nbsp;Xinyu Li ,&nbsp;Liang Gao","doi":"10.1016/j.rcim.2025.103029","DOIUrl":"10.1016/j.rcim.2025.103029","url":null,"abstract":"<div><div>The growing demand for flexible production systems is driven by product diversity and fluctuating order volumes. Seasonal variations can lead to imbalances between available machines and order demands, making efficient resource configuration critical before production begins. This paper addresses the optimization of machine configuration and scheduling in the hybrid flow shop, incorporating the order delivery time window. Most previous studies have focused on fixed machine numbers, while this study considers uncertain machine availability. This paper presents the mixed-integer linear programming model of the problem, and a linear programming (LP)-driven variable strategies evolutionary approach. The proposed approach combines an evolutionary algorithm with LP-driven neighborhood search for sequence optimization. Three strategies are constructed by narrowing down the search scope, which effectively reduces the algorithm's stagnation time and speeds up the convergence. To evaluate the effectiveness of the approach, the scales of application of the three LP strategies are first tested. Then ablation and comparison experiments are conducted, which show that the proposed approach generally agrees with the MILP results on small-scale problems and with smaller time resources. Experiments on 60 sets of large-scale problems show that the proposed approach has significant advantages over 5 state-of-the-art evolutionary algorithms and the MILP model. Additionally, the experiments show that the LP-driven strategy can improve the algorithm efficiency by about 13.32 % compared with the conventional strategy. These results demonstrate the potential of the LP-driven evolutionary approach for solving the complex scheduling problem.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103029"},"PeriodicalIF":9.1,"publicationDate":"2025-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143851451","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 novel impact-based dynamic motion planning of parallel kinematic forming robot under heavy load","authors":"Fangyan Zheng,&nbsp;Yi Zhong,&nbsp;Xinghui Han,&nbsp;Lin Hua,&nbsp;Shuai Xin","doi":"10.1016/j.rcim.2025.103031","DOIUrl":"10.1016/j.rcim.2025.103031","url":null,"abstract":"<div><div>Dynamic accuracy is significant for industrial robot in application. To realize this, many methods such as increasing damping, changing materials, and optimizing structures are developed and applied to the robots with relative low load (thousands of Newtons). However, for the application of heavy load (millions of Newtons), the dynamic error generation mechanism is different, these methods are neither suitable nor economic. To address this issue, the dynamic error generation mechanism of industrial robots under heavy load is revealed and a novel impact-based motion planning method is proposed.</div><div>Take parallel kinematic forming robot (PKFR) with load of 6MN as an example, the rigid-flexible coupling dynamic model considering joint clearance is first established and experimentally validated by a 70 % prediction accuracy. The dynamic error reaches up to 3.26 mm in position and 5.5mrad in angle. The impact forces are up to 10–20 times of driving force and it occur 12 times in a working cycle when one of the driving force approaches to zero. Further, the dynamic error generation mechanism is revealed, namely dynamic error of platform is mainly generated by the vibration impact sourced from the high joint clearance and the high variation of drive velocity. Thus, a novel impact-based dynamic motion planning method is proposed through reduction of the slider velocity at moment of impact. Using this method, the dynamic error is greatly reduced (42.82 % of position error and 34.82 % of angular error) in theory. Finally, an aircraft window frame is formed, showing a 36.53 % reduction in outer thickness error and a 33.65 % reduction in inner thickness error by using the proposed method. This method provides a new approach to reduce the dynamic error of industrial robots under heavy load and has high application potential due to its economic benefits without change of mechanical system.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103031"},"PeriodicalIF":9.1,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143833891","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":"Fast on-site assembly mating surface status analysis based on Skin Model Shapes","authors":"Li Jinyue ,&nbsp;Zhao Gang ,&nbsp;Xu Maocheng ,&nbsp;Xu Pengpeng ,&nbsp;Zhang Pengfei","doi":"10.1016/j.rcim.2025.103024","DOIUrl":"10.1016/j.rcim.2025.103024","url":null,"abstract":"<div><div>Surface morphology has a considerable impact on the assembly of high-precision mechanical products. To address this issue, an adapted Skin Model Shapes modeling method for on-site assembly is employed. Meanwhile, the computational time would drastically expand as the exponential growth of measurement data hlfaced the increase in workpiece size. Thus, a fast surface mating status analysis method is further introduced for point cloud based Skin Model Shapes. Computational efficiency is improved by reducing the search points through region division. The proposed method is applied to the assembly of rotary components in aero-engines. Test workpieces with detailed morphology and measuring devices are developed to conduct assembly experiments. The experimental results show that the contact areas predicted by the simulation results overlap more than 70% with respect to the actual results, and the computational time is reduced by 82% compared with the existing method.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103024"},"PeriodicalIF":9.1,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143833892","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":"Towards human-centric manufacturing: A reinforcement learning method for physical exertion alleviation in HRCA","authors":"Yingchao You,&nbsp;Ze Ji","doi":"10.1016/j.rcim.2025.103025","DOIUrl":"10.1016/j.rcim.2025.103025","url":null,"abstract":"<div><div>The advancement of the manufacturing system towards more human-centric, emphasising not only efficiency but also the well-being of workers. However, task planning in human–robot collaborative assembly (HRCA) remains challenging, when considering the physical exertion alleviation of workers, due to the complexities of physical exertion estimation and variations in human assembly operations. Different from conventional methods, this paper proposes a task planning method for physical exertion alleviation of workers in HRCA by leveraging the reinforcement learning (RL) method to train a policy. Initially, a musculoskeletal model-based method driven by human movement data to assess workers’ physical exertion is integrated into this work. Then, the policy is trained in a DuelingDQN-AM framework, utilising a carefully designed reward function informed by the estimated physical exertion of workers. The effectiveness of this approach has been validated through a simulation experiment and a proof-of-concept real assembly experiment. Simulation experiment results demonstrate the advantages of DuelingDQN-AM over other methods in terms of convergence speed and stability across multiple cycles and products of varying complexity. Additionally, real-world experiment results show that the RL strategy reduced physical exertion by 15.63<span><math><mtext>%</mtext></math></span> compared to the baseline method.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103025"},"PeriodicalIF":9.1,"publicationDate":"2025-04-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143834281","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":"Teaching robots to weld by leveraging human expertise","authors":"Junfu Zhou ,&nbsp;Abdelkhalick Mohammad ,&nbsp;Tianyi Zeng ,&nbsp;Dragos Axinte ,&nbsp;Iain Wright ,&nbsp;Richard March","doi":"10.1016/j.rcim.2025.103027","DOIUrl":"10.1016/j.rcim.2025.103027","url":null,"abstract":"<div><div>Robotic welding systems are pivotal in various manufacturing sectors, such as aerospace, construction, automotive, and maritime industries, due to their ability to operate in challenging environments with fewer physical constraints compared to human welders. However, their lack of process knowledge and adaptability necessitates heavy reliance on experienced technicians for process planning. To mitigate these challenges, a novel robotic welding system is proposed, focusing on learning from manual operations. In the proposed approach, proficient welders execute basic tasks, such as welding simple lines or arcs, while their actions are recorded using an operation tracking system. Then key welding parameters, such as torch travelling speed, welding arc length, welding angle, welding current, and wire feeding rate, are extracted and stored in a skill library. New welding tasks are segmented into the elements of the library. These are matched with archived parameters to plan the process for the robotic welding system, effectively transferring welding expertise to the automated system. Experiments have been conducted to verify the system. A skilled welder was asked to weld linear and arc-shaped grooves on stainless steel workpieces, while the welder’s skills were tracked, extracted, and stored digitally. These skills were further used to plan the robotic welding system to execute new complex tasks, such as polynomial curves. Welding results from the robot show a quality that is on par with that of a skilled welder.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103027"},"PeriodicalIF":9.1,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143816630","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":"Physics-based modeling and intelligent optimal decision method for digital twin system towards sustainable CNC equipment","authors":"Shulong Mei ,&nbsp;Yang Xie ,&nbsp;Jinfeng Liu ,&nbsp;Jianzhao Wu ,&nbsp;Chaoyong Zhang","doi":"10.1016/j.rcim.2025.103028","DOIUrl":"10.1016/j.rcim.2025.103028","url":null,"abstract":"<div><div>With the continuous advancement of smart manufacturing technologies, traditional machine tool production is evolving toward greater integration and intelligence, particularly in addressing multi-objective optimization challenges such as energy efficiency, operational effectiveness, and noise reduction. To optimize machine tool performance under dynamically changing milling process parameters, a multi-objective optimization strategy for milling parameters based on digital twin technology is proposed. A virtual machine model is developed using the milling process parameters and data from the physical machine tool. Real-time data interaction between the physical and virtual machines is facilitated through edge gateways and the industrial internet, dynamically updating the motion relationships within the geometric model to establish a digital twin of the machine tool. An initial analysis of energy consumption characteristics in the milling process is conducted, followed by the construction of a multi-objective optimization model that incorporates interactions between the physical and virtual machine tools through digital twin technology. Through data exchange between the physical and virtual models, real-time operational data from the machine tool are gathered. The Optuna-optimized XGBoost algorithm (Optuna-XGBoost) is applied for target prediction, combined with an improved multi-objective rime optimization algorithm (IMORIME) to optimize the milling process. Finally, the TOPSIS decision analysis method is employed to evaluate the Pareto solution set, identifying the optimal combination of process parameters. Experimental results demonstrate that the digital twin-based optimization approach achieves significant reductions in spindle energy consumption by 11.96 %, specific cutting energy by 28.24 %, and noise levels by 11.38 % compared to traditional methods, while also enhancing the visualization of machine tool information.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103028"},"PeriodicalIF":9.1,"publicationDate":"2025-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143816246","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":"Proposal of a complexity model for human-robot collaboration assembly processes","authors":"Matteo Capponi,&nbsp;Riccardo Gervasi,&nbsp;Luca Mastrogiacomo,&nbsp;Fiorenzo Franceschini","doi":"10.1016/j.rcim.2025.103026","DOIUrl":"10.1016/j.rcim.2025.103026","url":null,"abstract":"<div><div>Assembly complexity in manual processes has been widely addressed over the years in manufacturing-related literature. The concept of complexity indeed is linked to the cognitive and physical effort required on behalf of the human operator in completing the assembly process and is directly linked to the occurrence of process failures and inefficiencies. In the light of the introduction of novel technologies such as collaborative robotics such paradigm should be revised. This paper presents a proposal for a complexity model, i.e., “C<img>HRC model”, for Human-Robot Collaboration assembly processes. C<img>HRC model provides a multidimensional framework and a practical tool for analysing the complexity of collaborative assembly processes performed by humans supported by collaborative robots. In this situation, the collaboration with the robot may require an additional effort from the human operator, resulting in a more complex activity and thus more error prone. In this regard, the C<img>HRC model integrates insights from multiple disciplines to provide an overview of collaborative assembly complexity based on four layers: product complexity, assembly complexity, interaction complexity and collaboration complexity. The conceptual foundation of the C<img>HRC model is thoroughly detailed and supported by a review of the relevant literature. Hence, the paper uses the complexity formulation proposed by Samy and ElMaraghy as a basis to provide a quantitative approach. The model is then applied to practical case studies to demonstrate its application and illustrate how it can enhance the understanding of effective human-robot collaboration. This provides process designers with a practical tool to support design and improve collaborative assembly processes.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103026"},"PeriodicalIF":9.1,"publicationDate":"2025-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143777255","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Novel Point Cloud-Driven Framework for Enhanced Multi-Views Model Reconstruction and Robotic Arc Welding Trajectory Generation","authors":"Kai Liu ,&nbsp;Yang Lu ,&nbsp;Mingzhen Rao ,&nbsp;Zhongxi Sheng ,&nbsp;Wei Zhang ,&nbsp;Zhengbin Zhong ,&nbsp;Xiao Yang ,&nbsp;Runquan Xiao ,&nbsp;Huabin Chen","doi":"10.1016/j.rcim.2025.103022","DOIUrl":"10.1016/j.rcim.2025.103022","url":null,"abstract":"<div><div>With the rapid advancement of robotics technology, robotic welding has become essential for improving the welding efficiency of large-scale complex components, while reducing the workload of welders. However, the large size and intricate weld structures present significant challenges in obtaining comprehensive point cloud and accurate welding paths, which further hinders the development and application of robotic welding in this field. To address these problems, this article proposes a novel point cloud-driven framework for enhanced multi-views model reconstruction and robotic arc welding trajectory generation of large-scale complex components. To determine the corresponding point pairs and optimal transformation used for point cloud alignment, we introduce an improved bidirectional nearest neighbor (IBNN) algorithm combined with a Levenberg-Marquardt iterative closest point (LM-ICP) approach, which enables precise and fast stitching of multi-views point clouds. We further propose an edge intensity (EI) response algorithm for efficient extraction of weld seams feature points from the point cloud, followed by B-spline curve fitting to generate smooth and accurate welding trajectories. Additionally, the welding torch pose is estimated by integrating the weld seams region (WSR) point cloud with the welding trajectories, enabling the robot to perform autonomous welding in the correct orientation. Experimental results show that the proposed framework outperforms traditional methods in both accuracy and efficiency, with a maximum error (ME) of about 0.6 mm, a root mean square error (RMSE) of approximately 0.4 mm, and a running time of around 7 s, which has a certain industrial application value.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103022"},"PeriodicalIF":9.1,"publicationDate":"2025-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143769201","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}