Robotics and Computer-integrated Manufacturing最新文献_第3页

A digital-twin framework for assembling of cylindrical parts 用于圆柱形零件装配的数字孪生框架

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-21 DOI: 10.1016/j.rcim.2026.103240

Yimin Song , Chen Li , Binbin Lian , Qi Li , Tao Sun

Digital twin (DT) has been recognized as a promising technology for enhanced planning, monitoring and control of automatic assembly, with the capability of efficiency, adaptability and flexibility. While most of DT-based assembly focus on electronic products, limited attention has been paid to the assembling of heavy and large-size product with tight tolerance. This paper presents a DT model facilitating prediction and real-time adjustment for intelligent assembling of cylindrical parts. Vision guided feature fitting and coordinate frame construction are presented. Herein, assembly targets are defined considering eight sets of pin and hole, and the contacting planes. This approach improves the assembly success rate. To ensure efficient and robust robot assembly, we proposed a prediction model based on the DT system. Unknown errors and uncertainty of physical space are considered by small displacement torsor (SDT) theory and Monte Carlo simulation (MCS). Assembly planning and execution would efficiently adjust guided by the prediction result. A middle point is set in the robot planning that leaves pure translation in the docking phase. Real-time adjustment method is proposed to accurately assemble the cylindrical parts. Simulations and experiments are carried out to verify the effectiveness and feasibility of the proposed DT-based assembling method. The results show that the prediction results are the same as the actual assembly. Our assembly strategy achieves 97.31% success rate. By employing the assembly strategy and real-time adjustment, our method ensures that the majority of axes mismatch is below 0.1mm/0.05 deg, plane non-contacting below 0.05 mm.

数字孪生技术（Digital twin， DT）以其高效、适应性强、灵活的特点，被认为是一种很有前途的自动化装配规划、监测和控制技术。目前，基于3d打印技术的装配大多集中在电子产品上，而对重、大尺寸、公差要求高的产品的装配关注甚少。针对圆柱件智能装配，提出了一种便于预测和实时调整的DT模型。提出了视觉引导特征拟合和坐标框架构建方法。其中，考虑8组销孔和接触面，定义装配目标。这种方法提高了装配成功率。为了保证机器人装配的效率和鲁棒性，提出了一种基于DT系统的预测模型。小位移变形量（SDT）理论和蒙特卡罗模拟（MCS）考虑了物理空间的未知误差和不确定性。在预测结果的指导下，有效地调整装配规划和执行。在机器人规划中设置一个中间点，在对接阶段保留纯平移。提出了圆柱零件精确装配的实时调整方法。仿真和实验验证了该方法的有效性和可行性。结果表明，预测结果与实际装配相吻合。我们的装配策略达到97.31%的成功率。通过采用装配策略和实时调整，保证了大部分轴错配在0.1mm/0.05°以下，平面不接触在0.05 mm以下。

{"title":"A digital-twin framework for assembling of cylindrical parts","authors":"Yimin Song , Chen Li , Binbin Lian , Qi Li , Tao Sun","doi":"10.1016/j.rcim.2026.103240","DOIUrl":"10.1016/j.rcim.2026.103240","url":null,"abstract":"<div><div>Digital twin (DT) has been recognized as a promising technology for enhanced planning, monitoring and control of automatic assembly, with the capability of efficiency, adaptability and flexibility. While most of DT-based assembly focus on electronic products, limited attention has been paid to the assembling of heavy and large-size product with tight tolerance. This paper presents a DT model facilitating prediction and real-time adjustment for intelligent assembling of cylindrical parts. Vision guided feature fitting and coordinate frame construction are presented. Herein, assembly targets are defined considering eight sets of pin and hole, and the contacting planes. This approach improves the assembly success rate. To ensure efficient and robust robot assembly, we proposed a prediction model based on the DT system. Unknown errors and uncertainty of physical space are considered by small displacement torsor (SDT) theory and Monte Carlo simulation (MCS). Assembly planning and execution would efficiently adjust guided by the prediction result. A middle point is set in the robot planning that leaves pure translation in the docking phase. Real-time adjustment method is proposed to accurately assemble the cylindrical parts. Simulations and experiments are carried out to verify the effectiveness and feasibility of the proposed DT-based assembling method. The results show that the prediction results are the same as the actual assembly. Our assembly strategy achieves 97.31% success rate. By employing the assembly strategy and real-time adjustment, our method ensures that the majority of axes mismatch is below 0.1mm/0.05 deg, plane non-contacting below 0.05 mm.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103240"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146014798","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}

引用次数: 0

Toward generalizable robotic assembly: A prior-guided deep reinforcement learning approach with multi-sensor information 面向一般化机器人装配：基于多传感器信息的先验引导深度强化学习方法

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-22 DOI: 10.1016/j.rcim.2026.103242

Zilu Zhu , Yongkui Liu , Qianji Wang , Zinan Wang , Lihui Wang , Sichao Liu , Bin Zi , Lin Zhang

The rise of personalized manufacturing presents significant challenges for robotic assembly. While learning-based methods offer promising solutions, they often suffer from low training efficiency and poor generalization. To address these limitations, this paper proposes an efficient prior-guided (PG) deep reinforcement learning (DRL) approach for generalizable robotic assembly using multi-sensor information. First, a phased multi-sensor information fusion method is introduced. Then, a visual feature extraction method that combines MobileNetV3-Lite with conventional digital image processing and a rule-based force feature extraction method are designed to extract lower-dimensional features as prior-guided knowledge. Based on the methods above, a Soft Actor-Critic (SAC) algorithm that integrates Gated Recurrent Unit (GRU) network architecture with PG is proposed, thereby enabling efficient assembly skill learning. Simulations and physical experiments with respect to three typical assembly skills, i.e., search, alignment, and insertion, are conducted. Results indicate that, compared with the baseline SAC algorithm, our feature extraction method reduces visual feature dimensions by 93.75% and provides accurate prior-guided knowledge for DRL. The proposed assembly skill learning algorithm achieves a 30.16% reduction in average training time and a 16.82% decrease in average completion step. Furthermore, all learned skills can be rapidly transferred across different objects, and all assembly tasks are completed efficiently and compliantly with an average success rate of 96.86%.

个性化制造的兴起对机器人装配提出了重大挑战。虽然基于学习的方法提供了很好的解决方案，但它们往往存在训练效率低和泛化能力差的问题。为了解决这些限制，本文提出了一种有效的先验引导（PG）深度强化学习（DRL）方法，用于使用多传感器信息的可泛化机器人装配。首先，介绍了一种分阶段的多传感器信息融合方法。然后，设计了MobileNetV3-Lite与传统数字图像处理相结合的视觉特征提取方法和基于规则的力特征提取方法，作为先验引导知识提取低维特征。在此基础上，提出了一种将门控循环单元（GRU）网络体系结构与PG相结合的软Actor-Critic （SAC）算法，从而实现了高效的装配技能学习。针对三种典型装配技能，即搜索、对准和插入，进行了仿真和物理实验。结果表明，与基线SAC算法相比，我们的特征提取方法将视觉特征维数降低了93.75%，为DRL提供了准确的先验引导知识。提出的装配技能学习算法平均训练时间减少30.16%，平均完成步长减少16.82%。此外，所有学习到的技能都可以在不同的对象之间快速转移，所有装配任务都能高效、合规地完成，平均成功率为96.86%。

{"title":"Toward generalizable robotic assembly: A prior-guided deep reinforcement learning approach with multi-sensor information","authors":"Zilu Zhu , Yongkui Liu , Qianji Wang , Zinan Wang , Lihui Wang , Sichao Liu , Bin Zi , Lin Zhang","doi":"10.1016/j.rcim.2026.103242","DOIUrl":"10.1016/j.rcim.2026.103242","url":null,"abstract":"<div><div>The rise of personalized manufacturing presents significant challenges for robotic assembly. While learning-based methods offer promising solutions, they often suffer from low training efficiency and poor generalization. To address these limitations, this paper proposes an efficient prior-guided (PG) deep reinforcement learning (DRL) approach for generalizable robotic assembly using multi-sensor information. First, a phased multi-sensor information fusion method is introduced. Then, a visual feature extraction method that combines MobileNetV3-Lite with conventional digital image processing and a rule-based force feature extraction method are designed to extract lower-dimensional features as prior-guided knowledge. Based on the methods above, a Soft Actor-Critic (SAC) algorithm that integrates Gated Recurrent Unit (GRU) network architecture with PG is proposed, thereby enabling efficient assembly skill learning. Simulations and physical experiments with respect to three typical assembly skills, i.e., search, alignment, and insertion, are conducted. Results indicate that, compared with the baseline SAC algorithm, our feature extraction method reduces visual feature dimensions by 93.75% and provides accurate prior-guided knowledge for DRL. The proposed assembly skill learning algorithm achieves a 30.16% reduction in average training time and a 16.82% decrease in average completion step. Furthermore, all learned skills can be rapidly transferred across different objects, and all assembly tasks are completed efficiently and compliantly with an average success rate of 96.86%.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103242"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146033276","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}

引用次数: 0

A digital twin-driven deviation analysis approach for aircraft hydraulic pipelines assembly 飞机液压管路装配双驱动偏差数字分析方法

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-10 DOI: 10.1016/j.rcim.2026.103234

Liang Chen , Yue Zhao , Aihua Su , Wenqiang Yang , Yu Zheng

The assembly deviation of aircraft hydraulic pipelines plays a critical role in ensuring flight stability. Nonetheless, the presence of multiple dynamic factors originating from various sources, which influence the final quality throughout the assembly process, poses significant challenges to conventional deviation analysis methods that rely on theoretical deviation calculation models with predetermined assembly parameters. To address these limitations, a digital twin (DT)-driven framework is proposed for the realtime assembly deviation analysis of aircraft hydraulic pipelines. First, a DT model encompassing the total factor information of the assembly process is constructed, comprising the resource, process and quality models. Then, a DT-driven assembly quality analysis method is presented. By integrating multidimensional data within a virtual environment and employing DT models to estimate assembly deviations in realtime, the quality of the entire pipeline system can be effectively analyzed. Finally, the efficacy of the proposed method is demonstrated through a case study involving the aircraft hydraulic pipeline assembly process.

飞机液压管路的装配偏差对飞机的飞行稳定性起着至关重要的作用。然而，在装配过程中，各种来源的多种动态因素会影响最终质量，这对传统的偏差分析方法提出了重大挑战，这些方法依赖于具有预定装配参数的理论偏差计算模型。针对这些局限性，提出了一种用于飞机液压管路装配偏差实时分析的数字孪生驱动框架。首先，构建了包含装配过程全要素信息的DT模型，包括资源模型、过程模型和质量模型。然后，提出了一种基于dt驱动的装配质量分析方法。通过集成虚拟环境中的多维数据，利用DT模型实时估计装配偏差，可以有效地分析整个管道系统的质量。最后，以飞机液压管路装配过程为例，验证了该方法的有效性。

{"title":"A digital twin-driven deviation analysis approach for aircraft hydraulic pipelines assembly","authors":"Liang Chen , Yue Zhao , Aihua Su , Wenqiang Yang , Yu Zheng","doi":"10.1016/j.rcim.2026.103234","DOIUrl":"10.1016/j.rcim.2026.103234","url":null,"abstract":"<div><div>The assembly deviation of aircraft hydraulic pipelines plays a critical role in ensuring flight stability. Nonetheless, the presence of multiple dynamic factors originating from various sources, which influence the final quality throughout the assembly process, poses significant challenges to conventional deviation analysis methods that rely on theoretical deviation calculation models with predetermined assembly parameters. To address these limitations, a digital twin (DT)-driven framework is proposed for the realtime assembly deviation analysis of aircraft hydraulic pipelines. First, a DT model encompassing the total factor information of the assembly process is constructed, comprising the resource, process and quality models. Then, a DT-driven assembly quality analysis method is presented. By integrating multidimensional data within a virtual environment and employing DT models to estimate assembly deviations in realtime, the quality of the entire pipeline system can be effectively analyzed. Finally, the efficacy of the proposed method is demonstrated through a case study involving the aircraft hydraulic pipeline assembly process.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103234"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940155","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}

引用次数: 0

BrepHGNet: A face-vertex interaction heterogeneous graph neural network for feature recognition BrepHGNet：一个用于特征识别的面-顶点交互异构图神经网络

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-28 DOI: 10.1016/j.rcim.2026.103251

Weibo Li , Jie Zhang , Jiazhen Pang , Dangdang Zheng

In the context of smart manufacturing and robotic technology integration, Boundary representation (B-rep) model feature recognition is a crucial technical bottleneck linking design intent to robotic automated manufacturing. As a core link bridging computer-aided design and robotic machining execution, it offers robots semantic analysis of manufacturable features. However, existing methods are limited to analyze face features in the B-rep model and neglect in-depth modeling of the topological-geometric associations among geometric elements. This makes it hard for robots to accurately perceive multi-features. To address these issues, this paper proposes a heterogeneous graph learning framework named BrepHGNet. Firstly, a face-vertex interaction heterogeneous graph descriptor for B-rep models is constructed. Two distinct types of nodes, namely face nodes and vertex nodes, along with two types of relationships, face-adjacency-face and vertex-subordination-face, are defined. This construction serves to retain the hierarchical topological structure and geometric information inherent in B-rep models. Secondly, a vertex global shape mapping approach is introduced. By computing the Euclidean distances from vertices to sampling points on other faces, this method captures the impact of vertices within complex geometric structures. Thirdly, a heterogeneous graph neural network is built. Node features are updated via message passing and aggregation mechanisms tailored to different relationships. Finally, comparative experiments conducted on the manufacturing and real word datasets demonstrate that the proposed face-vertex interaction heterogeneous graph can effectively capture the internal geometric-topological associations within B-rep models, providing a new technical pathway for B-rep model feature recognition.

在智能制造与机器人技术融合的背景下，边界表示（B-rep）模型特征识别是连接设计意图与机器人自动化制造的关键技术瓶颈。它是连接计算机辅助设计和机器人加工执行的核心环节，为机器人提供可制造特征的语义分析。然而，现有的方法仅限于分析B-rep模型中的人脸特征，而忽略了对几何元素之间拓扑-几何关联的深入建模。这使得机器人很难准确地感知多种特征。为了解决这些问题，本文提出了一个名为BrepHGNet的异构图学习框架。首先，构造了面向B-rep模型的面-顶点交互异构图描述符。定义了两种不同类型的节点，即面节点和顶点节点，以及两种类型的关系，即脸-邻接-脸和顶点-从属-脸。这种结构有助于保留B-rep模型中固有的分层拓扑结构和几何信息。其次，介绍了一种顶点全局形状映射方法。该方法通过计算顶点到其他面采样点的欧氏距离，捕捉复杂几何结构中顶点的影响。第三，构建了异构图神经网络。节点特性通过针对不同关系定制的消息传递和聚合机制进行更新。最后，在制造数据集和真实单词数据集上进行的对比实验表明，所提出的人脸-顶点交互异构图能够有效地捕获B-rep模型内部的几何拓扑关联，为B-rep模型特征识别提供了新的技术途径。

{"title":"BrepHGNet: A face-vertex interaction heterogeneous graph neural network for feature recognition","authors":"Weibo Li , Jie Zhang , Jiazhen Pang , Dangdang Zheng","doi":"10.1016/j.rcim.2026.103251","DOIUrl":"10.1016/j.rcim.2026.103251","url":null,"abstract":"<div><div>In the context of smart manufacturing and robotic technology integration, Boundary representation (B-rep) model feature recognition is a crucial technical bottleneck linking design intent to robotic automated manufacturing. As a core link bridging computer-aided design and robotic machining execution, it offers robots semantic analysis of manufacturable features. However, existing methods are limited to analyze face features in the B-rep model and neglect in-depth modeling of the topological-geometric associations among geometric elements. This makes it hard for robots to accurately perceive multi-features. To address these issues, this paper proposes a heterogeneous graph learning framework named BrepHGNet. Firstly, a face-vertex interaction heterogeneous graph descriptor for B-rep models is constructed. Two distinct types of nodes, namely face nodes and vertex nodes, along with two types of relationships, face-adjacency-face and vertex-subordination-face, are defined. This construction serves to retain the hierarchical topological structure and geometric information inherent in B-rep models. Secondly, a vertex global shape mapping approach is introduced. By computing the Euclidean distances from vertices to sampling points on other faces, this method captures the impact of vertices within complex geometric structures. Thirdly, a heterogeneous graph neural network is built. Node features are updated via message passing and aggregation mechanisms tailored to different relationships. Finally, comparative experiments conducted on the manufacturing and real word datasets demonstrate that the proposed face-vertex interaction heterogeneous graph can effectively capture the internal geometric-topological associations within B-rep models, providing a new technical pathway for B-rep model feature recognition.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103251"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072476","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}

引用次数: 0

Adaptive active decoding and novel disjunctive graph-based improved genetic algorithm for multi-type machine robot cell scheduling in mass customization 大规模定制中多类型机器机器人单元调度的自适应主动解码和基于析取图的改进遗传算法

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-23 DOI: 10.1016/j.rcim.2026.103246

Yue Teng , Tianhong Wang , Xinyu Li , Chunjiang Zhang , Liang Gao , Ziyue Wang , Weiming Shen

Mass customization represents a critical evolution in modern manufacturing. To achieve efficient large-scale production of low-volume and high-variety products, designing optimized robot cells for flexible automation has become a universal challenge for manufacturers. While our prior research has effectively addressed scheduling problem in robot cells with discrete processing machines (DPMs, each processing one job at a time), the integration of both DPMs and batch processing machines (BPMs, each process multiple jobs simultaneously) introduces significant complexity for fully utilizing productive capacities. This paper investigates the Multi-Type Machine Robot Cell Scheduling Problem (MRCSP) incorporating both DPMs and BPMs and the objective is to minimize makespan. Firstly, a mixed-integer linear programming (MILP) model is formulated to describe MRCSP exactly. Recognizing the challenge of converting batch-aware two-vector encoding into feasible schedules, an adaptive active decoding strategy termed selective insertion batch decoding (SIBD) is proposed. An improved genetic algorithm (IGA) is then developed integrating this tailored encoding/decoding approach and a novel disjunctive graph. Furthermore, a batch neighborhood structure (BN) leveraging problem-specific characteristics is designed. The proposed MILP and IGA were validated on three FJSP-BPM benchmarks. Computational results demonstrate that IGA outperforms existing methods across all instances. In real-world production case studies, the approach achieved a 15.02 % average makespan reduction compared to prior methods, significantly improving resource utilization at a robot cell in southern China.

大规模定制是现代制造业的一个重要演变。为了实现小批量、多品种产品的高效大规模生产，设计优化的柔性自动化机器人单元已成为制造商普遍面临的挑战。虽然我们之前的研究已经有效地解决了离散加工机器（dpm，每次处理一个作业）在机器人单元中的调度问题，但dpm和批处理机器（bpm，每个工序同时处理多个作业）的集成为充分利用生产能力引入了显著的复杂性。本文研究了包含dpm和bpm的多类型机器机器人单元调度问题（MRCSP），其目标是最小化完工时间。首先，建立了一个混合整数线性规划（MILP）模型来精确描述MRCSP。考虑到将批感知双矢量编码转换为可行调度的挑战，提出了一种自适应主动解码策略，即选择性插入批解码（SIBD）。然后开发了一种改进的遗传算法（IGA），将这种定制的编码/解码方法与一种新的析取图相结合。在此基础上，设计了一种利用问题特征的批邻域结构（BN）。提出的MILP和IGA在三个FJSP-BPM基准上进行了验证。计算结果表明，IGA在所有实例中都优于现有方法。在实际生产案例研究中，与之前的方法相比，该方法实现了15.02%的平均完工时间减少，显著提高了中国南方机器人工厂的资源利用率。

{"title":"Adaptive active decoding and novel disjunctive graph-based improved genetic algorithm for multi-type machine robot cell scheduling in mass customization","authors":"Yue Teng , Tianhong Wang , Xinyu Li , Chunjiang Zhang , Liang Gao , Ziyue Wang , Weiming Shen","doi":"10.1016/j.rcim.2026.103246","DOIUrl":"10.1016/j.rcim.2026.103246","url":null,"abstract":"<div><div>Mass customization represents a critical evolution in modern manufacturing. To achieve efficient large-scale production of low-volume and high-variety products, designing optimized robot cells for flexible automation has become a universal challenge for manufacturers. While our prior research has effectively addressed scheduling problem in robot cells with discrete processing machines (DPMs, each processing one job at a time), the integration of both DPMs and batch processing machines (BPMs, each process multiple jobs simultaneously) introduces significant complexity for fully utilizing productive capacities. This paper investigates the Multi-Type Machine Robot Cell Scheduling Problem (MRCSP) incorporating both DPMs and BPMs and the objective is to minimize makespan. Firstly, a mixed-integer linear programming (MILP) model is formulated to describe MRCSP exactly. Recognizing the challenge of converting batch-aware two-vector encoding into feasible schedules, an adaptive active decoding strategy termed selective insertion batch decoding (SIBD) is proposed. An improved genetic algorithm (IGA) is then developed integrating this tailored encoding/decoding approach and a novel disjunctive graph. Furthermore, a batch neighborhood structure (BN) leveraging problem-specific characteristics is designed. The proposed MILP and IGA were validated on three FJSP-BPM benchmarks. Computational results demonstrate that IGA outperforms existing methods across all instances. In real-world production case studies, the approach achieved a 15.02 % average makespan reduction compared to prior methods, significantly improving resource utilization at a robot cell in southern China.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103246"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146033272","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}

引用次数: 0

Learning-based robotic machining error prediction for high precision manufacturing 基于学习的高精度机器人加工误差预测

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-02 DOI: 10.1016/j.rcim.2025.103217

Chaoyue Niu , Bin Chen , Simon Fletcher , Peace Onawumi , Erdem Ozturk , Mahdi Mahfouf , Visakan Kadirkamanathan

High precision machining with robots is an open challenge. Achieving precision of dimensional and geometrical features with robotic machining would require compensation via feedback control which relies on accurate error prediction. Machining error prediction is a complex problem in high-precision manufacturing, where effective solutions must accurately estimate geometrical errors in different workpieces while minimizing quality inspection costs. It is also compounded by the need for real-time estimation for feedback control. This paper introduces a novel approach for predicting the quality of milled workpieces using low-cost, in-process signals and machine learning. The proposed method fuses internal machine controller commands—comprising end-effector trajectory coordinates and angular changes of six revolute joints in the robotic arm—with external laser tracker sensing signals that capture the real trajectory of the milling tool and predicts dimensional errors as would be obtained by a Coordinate Measuring Machine (CMM). To overcome the lack of knowledge of the dependence of the part dimensional error on the available signals, models with varying combinations of the sensors and the length of the time window of historical data for inclusion in the model were evaluated. In addition, five machine learning algorithms were selected, trained, evaluated and validated on data from two distinct workpieces and various spatial configurations. The best machine learning model achieved a sevenfold improvement in dimensional error prediction compared to solely using laser tracker data, with mean absolute error reduced from 0.0756 mm to 0.0097 mm. This study demonstrates the feasibility of using low-cost, in-process sensing signals to predict high-precision quality dimensional data that is normally measured by costly CMMs, enabling rapid part quality inspection and significant potential cost reduction.

用机器人进行高精度加工是一个公开的挑战。机器人加工要实现尺寸和几何特征的精度，需要通过反馈控制进行补偿，而反馈控制依赖于精确的误差预测。加工误差预测是高精度制造中的一个复杂问题，有效的解决方案必须准确估计不同工件的几何误差，同时最大限度地降低质量检测成本。它还与反馈控制的实时估计需求相结合。本文介绍了一种利用低成本、过程中信号和机器学习预测铣削工件质量的新方法。该方法将机器内部控制器指令（包括末端执行器轨迹坐标和机械臂中六个旋转关节的角度变化）与外部激光跟踪器传感信号融合在一起，这些信号捕获铣刀的真实轨迹并预测由坐标测量机（CMM）获得的尺寸误差。为了克服零件尺寸误差对可用信号依赖性的缺乏，对具有不同传感器组合的模型和历史数据的时间窗口长度进行了评估。此外，在两种不同工件和不同空间配置的数据上选择、训练、评估和验证了五种机器学习算法。与仅使用激光跟踪器数据相比，最好的机器学习模型在尺寸误差预测方面实现了7倍的改进，平均绝对误差从0.0756 mm减少到0.0097 mm。本研究证明了使用低成本的过程传感信号来预测通常由昂贵的三坐标测量机测量的高精度质量尺寸数据的可行性，从而实现快速零件质量检测并显着降低潜在成本。

{"title":"Learning-based robotic machining error prediction for high precision manufacturing","authors":"Chaoyue Niu , Bin Chen , Simon Fletcher , Peace Onawumi , Erdem Ozturk , Mahdi Mahfouf , Visakan Kadirkamanathan","doi":"10.1016/j.rcim.2025.103217","DOIUrl":"10.1016/j.rcim.2025.103217","url":null,"abstract":"<div><div>High precision machining with robots is an open challenge. Achieving precision of dimensional and geometrical features with robotic machining would require compensation via feedback control which relies on accurate error prediction. Machining error prediction is a complex problem in high-precision manufacturing, where effective solutions must accurately estimate geometrical errors in different workpieces while minimizing quality inspection costs. It is also compounded by the need for real-time estimation for feedback control. This paper introduces a novel approach for predicting the quality of milled workpieces using low-cost, in-process signals and machine learning. The proposed method fuses internal machine controller commands—comprising end-effector trajectory coordinates and angular changes of six revolute joints in the robotic arm—with external laser tracker sensing signals that capture the real trajectory of the milling tool and predicts dimensional errors as would be obtained by a Coordinate Measuring Machine (CMM). To overcome the lack of knowledge of the dependence of the part dimensional error on the available signals, models with varying combinations of the sensors and the length of the time window of historical data for inclusion in the model were evaluated. In addition, five machine learning algorithms were selected, trained, evaluated and validated on data from two distinct workpieces and various spatial configurations. The best machine learning model achieved a sevenfold improvement in dimensional error prediction compared to solely using laser tracker data, with mean absolute error reduced from 0.0756 mm to 0.0097 mm. This study demonstrates the feasibility of using low-cost, in-process sensing signals to predict high-precision quality dimensional data that is normally measured by costly CMMs, enabling rapid part quality inspection and significant potential cost reduction.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103217"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145886408","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}

引用次数: 0

Research progress in the dynamics of heavy-duty robots from the perspective of machining process 基于加工过程视角的重型机器人动力学研究进展

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-27 DOI: 10.1016/j.rcim.2026.103250

Zhongqun Li , Qunli Shen , Wenjing Wu , Hailu Fan

Heavy-duty industrial robots are increasingly applied in machining due to their large workspace and flexible posture adjustment capabilities. However, their inherent low stiffness makes them highly susceptible to chatter during machining, which significantly restricts their further development and application. Conducting high-precision dynamic modeling for chatter prediction and adopting active/ passive chatter suppression techniques are crucial to achieving chatter-free machining. This paper systematically reviews the global research progress in the dynamics of heavy-duty robotic machining systems. Firstly, it outlines the core technologies and typical applications of heavy-duty robots in machining. Secondly, it comprehensively compares various modeling and prediction methods for the dynamic characteristics of heavy-duty robot end-effectors. Thirdly, it deeply analyzes the two main chatter mechanisms in robotic machining—regenerative chatter and modal coupling chatter—and their corresponding analytical and prediction methods. Subsequently, several representative chatter suppression technologies are summarized. Finally, conclusions are drawn based on the above analysis, and key directions for future research are proposed. Through a comprehensive review and in-depth exploration of the dynamics research of heavy-duty robot machining, this paper aims to provide valuable references and guidance for scholars in related fields.

重型工业机器人由于具有较大的工作空间和灵活的姿态调节能力，在机械加工中得到越来越多的应用。然而，其固有的低刚度使其在加工过程中极易产生颤振，这极大地限制了其进一步的发展和应用。实现无颤振加工的关键是建立高精度的颤振预测动力学模型和采用主动/被动颤振抑制技术。本文系统地综述了国内外重型机器人加工系统动力学的研究进展。首先，概述了重型机器人在机械加工中的核心技术和典型应用。其次，对重型机器人末端执行器动态特性的各种建模与预测方法进行了综合比较。再次，深入分析了机器人加工中的两种主要颤振机制——再生颤振和模态耦合颤振，以及相应的分析和预测方法。总结了几种具有代表性的颤振抑制技术。最后，在上述分析的基础上得出结论，并提出了未来研究的重点方向。本文旨在通过对重型机器人加工动力学研究的全面回顾和深入探讨，为相关领域的学者提供有价值的参考和指导。

{"title":"Research progress in the dynamics of heavy-duty robots from the perspective of machining process","authors":"Zhongqun Li , Qunli Shen , Wenjing Wu , Hailu Fan","doi":"10.1016/j.rcim.2026.103250","DOIUrl":"10.1016/j.rcim.2026.103250","url":null,"abstract":"<div><div>Heavy-duty industrial robots are increasingly applied in machining due to their large workspace and flexible posture adjustment capabilities. However, their inherent low stiffness makes them highly susceptible to chatter during machining, which significantly restricts their further development and application. Conducting high-precision dynamic modeling for chatter prediction and adopting active/ passive chatter suppression techniques are crucial to achieving chatter-free machining. This paper systematically reviews the global research progress in the dynamics of heavy-duty robotic machining systems. Firstly, it outlines the core technologies and typical applications of heavy-duty robots in machining. Secondly, it comprehensively compares various modeling and prediction methods for the dynamic characteristics of heavy-duty robot end-effectors. Thirdly, it deeply analyzes the two main chatter mechanisms in robotic machining—regenerative chatter and modal coupling chatter—and their corresponding analytical and prediction methods. Subsequently, several representative chatter suppression technologies are summarized. Finally, conclusions are drawn based on the above analysis, and key directions for future research are proposed. Through a comprehensive review and in-depth exploration of the dynamics research of heavy-duty robot machining, this paper aims to provide valuable references and guidance for scholars in related fields.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103250"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072481","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}

引用次数: 0

Constraint-based redundancy resolution for nozzle orientation in 8-axis robot-assisted DED: A case on revolved components 基于约束的八轴机器人辅助DED喷嘴定位冗余度求解：以旋转部件为例

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-27 DOI: 10.1016/j.rcim.2026.103248

Jiale Wu , Qi Liu , Jiachen Ye , Kai Ren , Yanlong Cao

Process planning is critical for robot-assisted Directed Energy Deposition (DED) on curved component manufacturing, which involves various factors such as laser parameters, nozzle orientation, velocity strategy, and non-printing paths. For the DED process with coaxial powder feeding, an 8-axis robot system consisting of a 6-axis robotic arm and a 2-axis positioner has 3 functionally redundant degrees of freedom, including axial rotation of the nozzle, tilt and rotation of the positioner. This research aims to achieve the optimal nozzle orientation for deposition by utilizing the redundancy of the 8-axis robot system. A nozzle orientation evaluation metric is initially proposed to evaluate that the nozzle axis vector is opposite to the gravity and aligned with the build surface normal vector. Subsequently, robot trajectory planning strategies with different nozzle orientation constraints are designed for revolved blades. Finally, actual printing was performed based on numerical simulation. The developed theoretical optimal robot trajectory planning strategies have achieved geometric accuracy primarily within ±1 mm. The method can be adapted to the fabrication of more complex curved components by appropriately relaxing the nozzle orientation constraints.

工艺规划是机器人辅助定向能沉积（DED）曲面部件制造的关键，涉及激光参数、喷嘴方向、速度策略和非打印路径等多种因素。对于同轴给粉DED工艺，由6轴机械臂和2轴定位器组成的8轴机器人系统具有3个功能冗余自由度，包括喷嘴轴向旋转、定位器倾斜和旋转。本研究的目的是利用八轴机器人系统的冗余性来实现最佳的喷嘴沉积方向。初步提出了一种喷嘴方向评价度量，用于评价喷嘴轴矢量是否与重力方向相反，是否与构建面法矢量对齐。随后，针对旋转叶片，设计了不同喷嘴方向约束下的机器人轨迹规划策略。最后，在数值模拟的基础上进行了实际打印。所开发的理论最优机器人轨迹规划策略的几何精度基本在±1mm以内。通过适当放宽喷嘴方向约束，该方法可以适用于制造更复杂的弯曲部件。

{"title":"Constraint-based redundancy resolution for nozzle orientation in 8-axis robot-assisted DED: A case on revolved components","authors":"Jiale Wu , Qi Liu , Jiachen Ye , Kai Ren , Yanlong Cao","doi":"10.1016/j.rcim.2026.103248","DOIUrl":"10.1016/j.rcim.2026.103248","url":null,"abstract":"<div><div>Process planning is critical for robot-assisted Directed Energy Deposition (DED) on curved component manufacturing, which involves various factors such as laser parameters, nozzle orientation, velocity strategy, and non-printing paths. For the DED process with coaxial powder feeding, an 8-axis robot system consisting of a 6-axis robotic arm and a 2-axis positioner has 3 functionally redundant degrees of freedom, including axial rotation of the nozzle, tilt and rotation of the positioner. This research aims to achieve the optimal nozzle orientation for deposition by utilizing the redundancy of the 8-axis robot system. A nozzle orientation evaluation metric is initially proposed to evaluate that the nozzle axis vector is opposite to the gravity and aligned with the build surface normal vector. Subsequently, robot trajectory planning strategies with different nozzle orientation constraints are designed for revolved blades. Finally, actual printing was performed based on numerical simulation. The developed theoretical optimal robot trajectory planning strategies have achieved geometric accuracy primarily within ±1 mm. The method can be adapted to the fabrication of more complex curved components by appropriately relaxing the nozzle orientation constraints.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103248"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146048117","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}

引用次数: 0

Modal-aware vibration evaluation and suppression method for robotic milling 铣削机器人模态感知振动评估与抑制方法

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-03 DOI: 10.1016/j.rcim.2025.103220

Zhao-Yang Liao , Hai-Long Xie , Bing Li , Yu Peng , Xue-Feng Zhou

Robotic milling is prone to chatter instability due to the inherent low stiffness of industrial manipulators, which significantly compromises machining accuracy. Accurate prediction of milling stability requires knowledge of the tool tip Frequency Response Function (FRF), which is closely related to the tool tip dynamics and highly sensitive to the robot’s posture. Traditional methods for obtaining the FRF rely on extensive experimental measurements and struggle to capture large variations in dynamic parameters across complex toolpaths. To address this challenge, this work proposes a modal-aware strategy for robotic vibration evaluation and suppression. A novel prediction method is developed by integrating Generative Adversarial Networks (GAN) with Gaussian Process Regression (GPR), enabling efficient estimation of robot modal parameters across the entire workspace. The proposed framework demonstrates strong robustness to joint angle noise and posture variations, ensuring reliable dynamic prediction under real-world machining uncertainties. Based on the predicted modal data, a stability indicator is constructed, and a posture optimization strategy is introduced by leveraging the robot’s kinematic redundancy to enhance dynamic stability during milling. Both simulation and experimental results demonstrate the effectiveness of the proposed strategy in improving machining stability, reducing energy entropy, and enhancing surface quality. Compared with traditional GPR, the proposed model achieves an

R^{2}

improvement of approximately 10% in modal parameter prediction.

由于工业机械臂固有的低刚度，机器人铣削容易产生颤振不稳定，这严重影响了加工精度。铣削稳定性的准确预测需要了解刀尖频响函数（FRF），它与刀尖动力学密切相关，对机器人的姿态高度敏感。获取FRF的传统方法依赖于大量的实验测量，难以捕获复杂刀具路径上动态参数的大变化。为了解决这一挑战，本工作提出了一种用于机器人振动评估和抑制的模态感知策略。将生成对抗网络（GAN）与高斯过程回归（GPR）相结合，提出了一种新的预测方法，能够在整个工作空间内有效地估计机器人模态参数。该框架对关节角度噪声和姿态变化具有较强的鲁棒性，确保了在实际加工不确定性下的可靠动态预测。基于预测的模态数据，构造了稳定性指标，并利用机器人的运动冗余引入姿态优化策略，提高铣削过程的动态稳定性。仿真和实验结果均证明了该策略在提高加工稳定性、降低能量熵和提高表面质量方面的有效性。与传统探地雷达模型相比，该模型的模态参数预测R2提高了约10%。

{"title":"Modal-aware vibration evaluation and suppression method for robotic milling","authors":"Zhao-Yang Liao , Hai-Long Xie , Bing Li , Yu Peng , Xue-Feng Zhou","doi":"10.1016/j.rcim.2025.103220","DOIUrl":"10.1016/j.rcim.2025.103220","url":null,"abstract":"<div><div>Robotic milling is prone to chatter instability due to the inherent low stiffness of industrial manipulators, which significantly compromises machining accuracy. Accurate prediction of milling stability requires knowledge of the tool tip Frequency Response Function (FRF), which is closely related to the tool tip dynamics and highly sensitive to the robot’s posture. Traditional methods for obtaining the FRF rely on extensive experimental measurements and struggle to capture large variations in dynamic parameters across complex toolpaths. To address this challenge, this work proposes a modal-aware strategy for robotic vibration evaluation and suppression. A novel prediction method is developed by integrating Generative Adversarial Networks (GAN) with Gaussian Process Regression (GPR), enabling efficient estimation of robot modal parameters across the entire workspace. The proposed framework demonstrates strong robustness to joint angle noise and posture variations, ensuring reliable dynamic prediction under real-world machining uncertainties. Based on the predicted modal data, a stability indicator is constructed, and a posture optimization strategy is introduced by leveraging the robot’s kinematic redundancy to enhance dynamic stability during milling. Both simulation and experimental results demonstrate the effectiveness of the proposed strategy in improving machining stability, reducing energy entropy, and enhancing surface quality. Compared with traditional GPR, the proposed model achieves an <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> improvement of approximately 10% in modal parameter prediction.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103220"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145894582","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}

引用次数: 0

Novel screw fastening manipulator for small-batch production: Multi-phase operations and Bayesian torque estimation 用于小批量生产的新型螺钉紧固机械手：多相操作和贝叶斯扭矩估计

IF 11.4 1区计算机科学 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Robotics and Computer-integrated Manufacturing

Pub Date : 2026-08-01 Epub Date: 2026-01-08 DOI: 10.1016/j.rcim.2026.103230

Gaokun Shi , Zhijian Wang , Guodong Lu , Hassen Nigatu

Screw fastening is a ubiquitous yet automation-resistant task in machinery assembly, particularly in small-batch production where flexibility, cost, and space constraints limit the adoption of conventional robotic systems. Existing solutions often rely on multi-motor designs with complex mechanisms and controls, resulting in high costs, bulky structures, and limited adaptability to varying screw types or confined workspaces. This paper presents a novel screw fastening manipulator that addresses these challenges through mechanical simplification and functional integration. The proposed two-actuator design automates four essential operations – grasping, pressing, tightening, and releasing – within a compact and efficient form factor. A cable-driven reorientation mechanism enables precise screw alignment in restricted environments, while the screw fastening manipulator autonomously regulates the screw feed rate, enabling the robot arm to remain stationary during the tightening process. This decoupling reduces system complexity and ensures consistent screw insertion. Furthermore, a physics-based Bayesian inference model is employed for real-time torque estimation and autonomous phase detection without the need for additional sensors. This sensorless control approach enhances system robustness, reduces hardware dependencies, and ensures optimal torque application through probabilistic decision-making. Experimental validations using M2.5–M6 ISO metric screws – common in home appliance assembly – demonstrate the adaptability, precision, and suitability of the screw fastening manipulator for constrained, small-batch manufacturing environment.

螺钉紧固是机械装配中普遍存在的自动化任务，特别是在小批量生产中，灵活性、成本和空间限制限制了传统机器人系统的采用。现有的解决方案通常依赖于具有复杂机构和控制的多电机设计，导致成本高，结构笨重，对不同螺杆类型或受限工作空间的适应性有限。本文提出了一种新型的螺钉紧固机械手，通过机械简化和功能集成来解决这些问题。提出的两个致动器设计自动化四个基本操作-抓取，按压，收紧和释放-在一个紧凑和高效的形式因素。钢丝绳驱动的重定向机构可以在受限环境下实现螺钉精确对准，螺钉紧固机械手可以自主调节螺钉进给速率，使机械臂在紧固过程中保持静止。这种分离降低了系统的复杂性，并确保了螺杆插入的一致性。此外，在不需要额外传感器的情况下，采用基于物理的贝叶斯推理模型进行实时扭矩估计和自主相位检测。这种无传感器控制方法增强了系统的鲁棒性，减少了硬件依赖性，并通过概率决策确保了最佳扭矩应用。使用家用电器装配中常见的M2.5-M6 ISO公制螺钉进行实验验证，证明了螺钉紧固机械手在受限的小批量制造环境中的适应性、精度和适用性。

{"title":"Novel screw fastening manipulator for small-batch production: Multi-phase operations and Bayesian torque estimation","authors":"Gaokun Shi , Zhijian Wang , Guodong Lu , Hassen Nigatu","doi":"10.1016/j.rcim.2026.103230","DOIUrl":"10.1016/j.rcim.2026.103230","url":null,"abstract":"<div><div>Screw fastening is a ubiquitous yet automation-resistant task in machinery assembly, particularly in small-batch production where flexibility, cost, and space constraints limit the adoption of conventional robotic systems. Existing solutions often rely on multi-motor designs with complex mechanisms and controls, resulting in high costs, bulky structures, and limited adaptability to varying screw types or confined workspaces. This paper presents a novel screw fastening manipulator that addresses these challenges through mechanical simplification and functional integration. The proposed two-actuator design automates four essential operations – grasping, pressing, tightening, and releasing – within a compact and efficient form factor. A cable-driven reorientation mechanism enables precise screw alignment in restricted environments, while the screw fastening manipulator autonomously regulates the screw feed rate, enabling the robot arm to remain stationary during the tightening process. This decoupling reduces system complexity and ensures consistent screw insertion. Furthermore, a physics-based Bayesian inference model is employed for real-time torque estimation and autonomous phase detection without the need for additional sensors. This sensorless control approach enhances system robustness, reduces hardware dependencies, and ensures optimal torque application through probabilistic decision-making. Experimental validations using M2.5–M6 ISO metric screws – common in home appliance assembly – demonstrate the adaptability, precision, and suitability of the screw fastening manipulator for constrained, small-batch manufacturing environment.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"100 ","pages":"Article 103230"},"PeriodicalIF":11.4,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145940158","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}

引用次数: 0