Search

Intelligent assembly of large-scale, complex structures using an intelligent manufacturing platform represents the future development direction for industrial manufacturing. During large-scale structural assembly processes, several bottleneck problems occur in the existing auxiliary assembly technology. First, the traditional LiDAR-based assembly technology is often limited by the openness of the manufacturing environment, in which there are blind spots, and continuous online assembly adjustment thus cannot be realized. Second, for assembly of large structures, a single-station LiDAR system cannot achieve complete coverage, which means that a multi-station combination method must be used to acquire the complete three-dimensional data; many more data errors are caused by the transfer between stations than by the measurement accuracy of a single station, which means that the overall system's measurement and adjustment errors are increased greatly. Third, because of the large numbers of structural components contained in a large assembly, the accumulated errors may lead to assembly interference, but the LiDAR-assisted assembly process does not have a feedback perception capability, and thus assembly component loss can easily be caused when assembly interference occurs. Therefore, this paper proposes to combine an optical fiber sensor network with digital twin technology, which will allow the test data from the assembly entity state in the real world to be applied to the “twin” model in the virtual world and thus solve the problems with test openness and data transfer. The problem of station and perception feedback is also addressed and represents the main innovation of this work. The system uses an optical fiber sensor network as a flexible sensing medium to monitor the strain field distribution within a complex area in real time, and then completes real-time parameter adjustment of the virtual assembly based on the distributed data. Complex areas include areas that are laser-unreachable, areas with complex contact surfaces, and areas with large-scale bending deformations. An assembly condition monitoring system is designed based on the optical fiber sensor network, and an assembly condition monitoring algorithm based on multiple physical quantities is proposed. The feasibility of use of the optical fiber sensor network as the real-state parameter acquisition module for the digital twin intelligent assembly system is discussed. The offset of any position in the test area is calculated using the convolutional neural network of a residual module to provide the compensation parameters required for the virtual model of the assembly structure. In the model optimization parameter module, a correction data table is obtained through iterative learning of the algorithm to realize state prediction from the test data. The experiment simulates a large-scale structure assembly process, and performs virtual and real mapping for a variety of situations with different assembly errors to enable correction of the digital twin data stream for the assembly process through the optical fiber sensor network. In the plane strain field calibration experiment, the maximum error among the test values for this system is 0.032 ​mm, and the average error is 0.014 ​mm. The results show that use of visual calibration can correct the test error to within a very small range. This result is equally applicable to gradient curvature surfaces and freeform surfaces. Statistics show that the average measurement accuracy error for regular surfaces is better than 11.2%, and the average measurement accuracy error for irregular surfaces is better than 14.8%. During simulation of large-scale structure assembly experiments, the average position deviation accuracy is 0.043 ​mm, which is in line with the designed accuracy.

To improve the assembly accuracy of intelligent assembly process, a 3D surface reconstruction system based on laser scanning for assembly targets and assembly positions was designed. The light source uses a line laser to provide structured light, enabling rapid scanning of the assembly target surface on a single line. Calculate the spatial relationship between the assembly target and the assembly position through the calibration board. A position inversion optimization algorithm has been proposed. By completing the transformation of feature points from local coordinates to global coordinates. The energy distribution in the central area of the CCD image plane was simulated and analyzed under different baseline lengths, and the optimal baseline position of 35 mm was determined. The experiment is divided into two parts, and the reconstruction of the three-dimensional surface of the free-form surface target based on laser scanning was completed in the laboratory. The optimized target 3D point cloud data is evenly distributed, with almost no scattered points on the main two local surfaces, which well reflects the surface structure of the target. In the actual outdoor assembly experiment, the system cooperated with the assembly vehicle to achieve the correct assembly of the assembly. The reconstruction accuracy deviation of most test points is very small, with over 85% of test points having an error of less than 0.54 mm. The average deviation of these points before and after optimization is 0.67 mm. This system has higher application value in the field of large-scale intelligent assembly.

In the current intelligent manufacturing scenario, single-mode perception limits the accuracy and safety of the human-robot collaboration process, and cannot adapt to the assembly operation requirements of robots in complex industrial environments. To solve these problems, this paper proposes a Human-Robot Collaboration (HRC) framework for a dual-robot intelligent assembly system based on multimodal perception. This framework aims to provide multimodal information interaction for robot control through gesture perception, speech perception, human body perception, and visual perception. To verify the effectiveness of this framework, it was applied to the assembly scene of the integrally shrouded blade-rotor system. Experiments show that the multimodal perception dual-robot intelligent assembly system can better realize the HRC assembly of the integrally shrouded blade and improve the intelligence of the assembly process. It shows that the framework can promote robots to flexibly and accurately perceive information among people, robots, and objects to complete intelligent assembly. The multimodal information interaction effectively improves the operational robustness of the dual-robot intelligent assembly system in complex industrial environments and complex tasks. [ABSTRACT FROM AUTHOR]

The assembly and maintenance of products in the aviation industry constitute a crucial aspect of the product life cycle, with numerous tasks still reliant on manual operations. In order to solve the problem of narrow operation spaces and blind areas in the processes of manual assembly and maintenance, we proposed an augmented reality (AR) assistant guidance method specifically designed for such scenarios. By employing a multi-modality anti-occlusion tracking algorithm, pose data of assembly parts can be obtained, upon which AR guidance information is displayed. Additionally, we proposed an assembly step identification method to alleviate user interaction pressure. We developed an AR visualization assistant guidance system and designed and conducted a user evaluation experiment to measure the learnability, usability, and mental effort required. The results demonstrate that our method significantly enhances training efficiency by 128.77%, as well as improving assembly and maintenance efficiency by 29.53% and 27.27% compared with traditional methods. Moreover, it has significant advantages in learnability, usability, and mental effort, providing a feasible and effective resolution for addressing blind areas during assembly and maintenance within the aviation industry. [ABSTRACT FROM AUTHOR]

The assembly process of large-scale and non-standard industrial equipment poses significant challenges due to its inherent scale-related complexity and proneness to errors, making it difficult to ensure process cost, production cycle, and assembly accuracy. In response to the limitations of traditional ineffective production models, this paper aims to explore and propose a digital twin (DT)-based technology paradigm for the intelligent assembly of large-scale and non-standard industrial equipment, focusing on both the equipment structure and assembly process levels. The paradigm incorporates key technologies that facilitate the integration of virtual and physical information, including the establishment and updating of DT models for assembly structures using actual data, the assessment of structural assemblability based on DT models, the planning and simulation of assembly processes, and the implementation of virtual commissioning technology tailored to the actual assembly process. The effectiveness of the proposed paradigm is demonstrated through a case study involving the actual assembly of a large-scale aerodynamic experimental equipment. The results confirm its ability to provide valuable technical support for the design, evaluation, and optimization of industrial equipment assembly processes. By leveraging the DT-based methodological system proposed in this paper, significant improvements in the transparency and intelligence of industrial equipment production processes can be achieved. [ABSTRACT FROM AUTHOR]

Digital twin (DT) technology has become an effective means of industrial digital transformation due to its ability to realize virtual-real mapping, data fusion, etc. In recent years, the application of DT technology to the assembly process of complex mechanical products in multiple fields can effectively overcome the drawbacks of the current assembly method, which is a new strategy for achieving high-precision and high-efficiency intelligent assembly. However, there are few reviews of DT's intelligent assembly technology and application for complex mechanical products. Based on the composition of the DT assembly system, the paper systematically summarizes the critical technologies of DT assembly, including DT model representation of the assembly objects, DT model construction of the assembly process, and DT data collection/management. The application of the DT assembly model based on virtual and real fusion in online monitoring, quality prediction, process planning, real-time control, etc. is comprehensively reviewed. Finally, the challenges faced by DT intelligent assembly of complex mechanical products are described, and future research directions are predicted. [ABSTRACT FROM AUTHOR]

An intelligent assembly method was designed to realize the intelligent assembly of the profiled thermal battery pack and improve its assembly accuracy. Firstly, as the number and size of different monomer batteries vary, this paper takes the monomer thermal battery assembly as the object, with a common shape circle assembly screw arrangement and an established process model. Then, the assembly also has an improved differential evolution algorithm for assembly arrangement and process on the number, location, tightening of the screw assembly, torque, and the order of solutions. According to this scheme, the assembly and the flatness test were carried out. The results showed that the bottom plate of the assembly frame was "concave in the middle and warped around", and the flatness error was large. The scheme was optimized by numerical simulation analysis. After optimization, the average offset of the floor plane was 0.04 mm, and the offset accounted for 0.028% of the overall height; the maximum offset was 0.094 mm and the offset was reduced by 0.312%. [ABSTRACT FROM AUTHOR]

In clinical practice, the irregular shapes of traumas pose a significant challenge in rapidly manufacturing personalized scaffolds. To address these challenges, inspired by LEGO ^® bricks, this study proposed a novel concept of modular scaffolds and developed an innovative system based on machine vision for their rapid and intelligent assembly tailored to defect shapes. Trapezoidal interfaces effectively connect standardized bone units based on magnesium-doped silicate calcium, ensuring high stability of the modular scaffolds, with compressive strength up to 135 MPa and bending strength up to 17 MPa. Through self-developed defect recognition and reconstruction algorithms, defect recognition and personalized assembly schemes for bone scaffolds can be achieved autonomously. Modular scaffolds seamlessly integrate with surrounding bone tissue, promoting new bone growth, with no apparent differences compared to fully 3D printed integral scaffolds in the skull and femur repair experiments. In summary, the adoption of modular scaffolds not only integrates personalization and standardization but also satisfies the optimal treatment window.

The assembly and maintenance of products in the aviation industry constitute a crucial aspect of the product life cycle, with numerous tasks still reliant on manual operations. In order to solve the problem of narrow operation spaces and blind areas in the processes of manual assembly and maintenance, we proposed an augmented reality (AR) assistant guidance method specifically designed for such scenarios. By employing a multi-modality anti-occlusion tracking algorithm, pose data of assembly parts can be obtained, upon which AR guidance information is displayed. Additionally, we proposed an assembly step identification method to alleviate user interaction pressure. We developed an AR visualization assistant guidance system and designed and conducted a user evaluation experiment to measure the learnability, usability, and mental effort required. The results demonstrate that our method significantly enhances training efficiency by 128.77%, as well as improving assembly and maintenance efficiency by 29.53% and 27.27% compared with traditional methods. Moreover, it has significant advantages in learnability, usability, and mental effort, providing a feasible and effective resolution for addressing blind areas during assembly and maintenance within the aviation industry.

Assembly timing planning, which aims to solve the assembly action sequence and assembly part sequence with the shortest assembly time as the goal, is a necessary and critical step in intelligent assembly process planning. However, the current focus of assembly process planning is assembly sequence planning, whereas little research has been performed on assembly timing planning. A novel assembly timing planning method based on knowledge and mixed integer linear programming (MILP) is proposed in this paper. First, a knowledge base of the assembly process for timing planning is constructed using ontology. Then, based on the proposed strategy of dividing assembly timing planning into within-group planning and between-group planning, a MILP model of assembly timing planning for automatic assembly system is constructed. In addition, a software that realizes timing planning through human–machine collaboration is developed to verify and visualize the proposed timing planning method. The implementation is as follows: assembly action sentences are formed by searching the ontology keyword library, then timing knowledge for the action sequence and assembly sequence is established, and finally optimal assembly timing results are obtained after the calculation. Compared with the traditional serial assembly process, this method significantly reduces the assembly time, thereby improving the assembly efficiency, and the assembly schedule can be obtained automatically and quickly to guide the assembly process design. [ABSTRACT FROM AUTHOR]

The assembly process of large-scale and non-standard industrial equipment poses significant challenges due to its inherent scale-related complexity and proneness to errors, making it difficult to ensure process cost, production cycle, and assembly accuracy. In response to the limitations of traditional ineffective production models, this paper aims to explore and propose a digital twin (DT)-based technology paradigm for the intelligent assembly of large-scale and non-standard industrial equipment, focusing on both the equipment structure and assembly process levels. The paradigm incorporates key technologies that facilitate the integration of virtual and physical information, including the establishment and updating of DT models for assembly structures using actual data, the assessment of structural assemblability based on DT models, the planning and simulation of assembly processes, and the implementation of virtual commissioning technology tailored to the actual assembly process. The effectiveness of the proposed paradigm is demonstrated through a case study involving the actual assembly of a large-scale aerodynamic experimental equipment. The results confirm its ability to provide valuable technical support for the design, evaluation, and optimization of industrial equipment assembly processes. By leveraging the DT-based methodological system proposed in this paper, significant improvements in the transparency and intelligence of industrial equipment production processes can be achieved.

• Three methods are implemented to enhance and improve YOLO model. • Priori information in recursive feature layer effectively guide the model training. • The proposed ARSPP extract richer feature layers and improve the network fitting ability. • The feature map contains more abundant information by fusing contiguous pyramid feature maps. • Providing guidance and reference value for designing and improving other object detection models. Object detection is the most important part in intelligent assembly tasks, accurate and fast detection for different targets can complete positioning and assembly tasks more automatically and efficiently. In this paper, a feature enhancement object detection model based on YOLO is proposed. Firstly, the expression ability of feature layer is enhanced through RFP (Recursive Feature Pyramid) structure. The ARSPP (Atrous Residual Spatial Pyramid Pooling) is proposed to have a further enhancement for the feature layers output by the backbone network, it improves the recognition performance for multi-scale targets of model by using different size of dilated convolution and residual connection. Finally, the contiguous pyramid features are fused and enhanced through the attention mechanism, the results are used for the input of next recursive or predictive output. The model proposed in this paper effectively improves the detection accuracy of YOLO, it has 3% MAP improvement in PASCAL VOC dataset. The validity and accuracy of the model are verified in the robot intelligent assembly recognition task. [Display omitted] [ABSTRACT FROM AUTHOR]

An intelligent assembly method was designed to realize the intelligent assembly of the profiled thermal battery pack and improve its assembly accuracy. Firstly, as the number and size of different monomer batteries vary, this paper takes the monomer thermal battery assembly as the object, with a common shape circle assembly screw arrangement and an established process model. Then, the assembly also has an improved differential evolution algorithm for assembly arrangement and process on the number, location, tightening of the screw assembly, torque, and the order of solutions. According to this scheme, the assembly and the flatness test were carried out. The results showed that the bottom plate of the assembly frame was “concave in the middle and warped around”, and the flatness error was large. The scheme was optimized by numerical simulation analysis. After optimization, the average offset of the floor plane was 0.04 mm, and the offset accounted for 0.028% of the overall height; the maximum offset was 0.094 mm and the offset was reduced by 0.312%.

Reasonable automatic guided vehicle path planning can shorten the transportation time of materials and improve the production efficiency of the intelligent assembly workshop. Ant colony algorithm is a widely used path planning method, however, it suffers from the shortcomings that being easy to fall into local optimum and low search efficiency. To overcome these shortcomings, first, this paper proposes a step optimization method to improve the search efficiency of the ant colony algorithm, and a path simplification method to avoid getting blindly tortuous paths; Second, to overcome the problem that the ant colony algorithm is easy to fall into the local optimum, this paper proposes an adaptive pheromone volatilization coefficient strategy, which uses different pheromone volatilization coefficients at different stages of the search path; third, for the path conflict problem of multiple automatic guided vehicles, this paper proposes a load balancing strategy to avoid it, which is based on the consideration that, path conflicts are caused by excessive concentration of multiple automatic guided vehicles paths. Extensive simulation results demonstrate the feasibility and efficiency of the proposed methods.

Designing an intelligent and autonomous system remains a great challenge in the assembly field. Most reinforcement learning (RL) methods are applied to experiments with relatively small state spaces. However, the complicated situation and high-dimensional spaces of the assembly environment cause traditional RL methods to behave poorly in terms of their efficiency and accuracy. In this paper, a model-driven adaptive proximal proximity optimization (MAPPO) method was presented to make the assembly system autonomously rectify the bolt posture error. In the MAPPO method, a probabilistic tree and adaptive reward mechanism were used to improve the calculation efficiency and accuracy of the traditional PPO method. The size of the action space was reduced by establishing a hierarchical logical relationship for each parameter with a probabilistic tree. Based on an adaptive reward mechanism, the phenomenon that the algorithm easily falls into local minima could be improved. Finally, the proposed method was verified based on the Unity simulation engine. The advancement and robustness of the proposed model were also validated by comparing different cases in simulations and experiments. The results revealed that MAPPO has better algorithm efficiency and accuracy compared with other state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

In the framework of an intelligent assembly information system, the three-dimensional model of a product serves as a critical information source for the system, which also provides initial product feature information for assembly sequence planning. At present, in the process of extracting information from the STEP neutral file, the products are rarely performed at the assembly level. In addition, the characteristic of semantic expression of the same parts included in a product is always ignored. In this paper, based on the theory of ontology, the OntoSTEP model is constructed to extract information from STEP neutral files at the assembly level. A set of extraction rules for the basic corresponding information of the same parts in the assembly are developed. Further, a SWRL rule library is established in Protégé to realize the comprehensive information extraction of the assembly hierarchy for a product from its 3D model, which lays the foundation for the assembly sequence planning. [ABSTRACT FROM AUTHOR]

In today's manufacturing environment, the technology of intelligent assembly has developed rapidly, and the technique level of intelligent assembly sequence planning has been improved significantly. However, the current assembly sequence planning always focuses on modern heuristic algorithms with single goals. Although these methods increase the automation level, they still cannot be applied to actual production because they ignore artificial experience and mature knowledge, which are of great value. This paper comprehensively considers the five key factors that affect the assembly sequence, proposes an assembly sequence planning method based on experience and knowledge, and realizes the sequence automatic reasoning by constructing the corresponding domain ontology. In addition, the OntoSTEP tool is used to complete the ontology mapping of geometric model's information, so as to verify the feasibility of the assembly sequence. The method has guiding significance for actual production, such as assisting process designer do assembly sequence planning effectively. [ABSTRACT FROM AUTHOR]

With the continuous improvement of function diversity, quality and reliability, the complexity and scale of significant products such as aerospace equipment are increasing. This trend poses severe challenges to the assembly process. The large scale of products requires the fixed-position assembly with multiple roles of workers in parallel, and it increases the difficulty of multi-person assembly. The high complexity of products makes the assembly process tedious. It is time-consuming and laborious for workers to understand the 2D documents, and it easily leads to mis-operation. The assembly process of complex products can produce huge amounts of data, which have the potential for improving efficiency and quality but remain underutilized. The high reliability of products puts forward high requirements for real-time detection of the assembly process. With the development of information technology, data driven approach and Augmented Reality (AR) provide an efficient way to change the above situation. Therefore, this paper proposes an intelligent collaboration assembly system (ICAS) combining real-time data driven method with AR technology. In the ICAS, real-time data driven method is used to realize the human-to-human and human-to-machine collaboration under the fixed-position assembly, and AR is introduced to the system to assist the assembly of large-scale complex products. The intelligent detection method in the ICAS combines 3D reconstruction with sensors measurement data, and the detection result is displayed in an intuitive way by AR. The proposed system is verified in the assembly process of a large-scale space deployable mechanism, and it paves the way for the further development of intelligent assembly in the future. [ABSTRACT FROM AUTHOR]

The automatic completion of multiple peg-in-hole assembly tasks by robots remains a formidable challenge because the traditional control strategies require a complex analysis of the contact model. In this paper, the assembly task is formulated as a Markov decision process, and a model-driven deep deterministic policy gradient algorithm is proposed to accomplish the assembly task through the learned policy without analyzing the contact states. In our algorithm, the learning process is driven by a simple traditional force controller. In addition, a feedback exploration strategy is proposed to ensure that our algorithm can efficiently explore the optimal assembly policy and avoid risky actions, which can address the data efficiency and guarantee stability in realistic assembly scenarios. To improve the learning efficiency, we utilize a fuzzy reward system for the complex assembly process. Then, simulations and realistic experiments of a dual peg-in-hole assembly demonstrate the effectiveness of the proposed algorithm. The advantages of the fuzzy reward system and feedback exploration strategy are validated by comparing the performances of different cases in simulations and experiments. [ABSTRACT FROM AUTHOR]

The widespread use of composite materials in the aviation industry result in the control of the assembly gaps between components a key topic. In this work, a gap optimization method based on component posture alignment is presented to assist in the intelligent assembly of aircraft composite panels. To quantify the assembly gap, a static and posture-dependent dynamic model is developed with the idea of discretization, and then a mathematical evaluation method for the assembly quality is proposed based on two new concepts: gap distribution coefficient and gap distribution factor. Moreover, with a multi-objective optimization method, the assembly gap is minimized and the gap distribution is optimized under the constraints of DOF and assembly tolerances, which provides new ideas for intelligent assembly of big composite panels. Finally, the proposed method is studied by numerical simulations of an assembly work in point cloud, as well as hardware experiments of a composite panel assembly work. These studies, taken together, demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Base of the complex automatic production is automation of technological process control, automatic transport, manipulation, feeding, workpiece and tool exchange. Industrial production develops, the new generation production systems -- intelligent production systems. As the progress in engineering production never stops, the use of computers, classical automation devices and flexible production systems will not suffice any longer, and it is now effective to contemplate a possible introduction of production systems of a new generation, so called intelligent production systems. All subsystems of such intelligent systems contain elements of so called machine intelligence (sensor elements). The introduction of such systems with machine-intelligence elements should completely remove humans from the production process. [ABSTRACT FROM AUTHOR]

A hidden Markov model (HMM)-based assembly contact state recognition system is designed and implemented. The system utilizes the force/torque data captured from a wrist force sensor to extract the intrinsic spatial relationships of contact formations arise from robotic assembly. The paper introduces a theoretical framework for contact recognition based on discrete HMM with special emphasis on the practical realization of the system using an industrial robot. With the detailed exposition of the major algorithms for solving the three key problems in HMM, namely, the evaluation of observation sequence probability, optimization of state sequence probability, and optimization of model parameters, a working prototype of HMM-based system is developed. The performance of the contact recognition system is investigated by using experimental studies. The results obtained clearly demonstrate that HMM is an effective means for assembly contact modelling and identification, and that the framework reported in this paper provides an essential ground work for the development of a practical intelligent robotic system. [Copyright &y& Elsevier]

To accelerate the commercial implementation of high-energy batteries, recent research thrusts have turned to the practicality of Si-based electrodes. Although numerous nanostructured Si-based materials with exceptional performance have been reported in the past 20 years, the practical development of high-energy Si-based batteries has been beset by the bias between industrial application with gravimetrical energy shortages and scientific research with volumetric limits. In this context, the microscale design of Si-based anodes with densified microstructure has been deemed as an impactful solution to tackle these critical issues. However, their large-scale application is plagued by inadequate cycling stability. In this review, we present the challenges in Si-based materials design and draw a realistic picture regarding practical electrode engineering. Critical appraisals of recent advances in microscale design of stable Si-based materials are presented, including interfacial tailoring of Si microscale electrode, surface modification of SiO x microscale electrode, and structural engineering of hierarchical microscale electrode. Thereafter, other practical metrics beyond active material are also explored, such as robust binder design, electrolyte exploration, prelithiation technology, and thick-electrode engineering. Finally, we provide a roadmap starting with material design and ending with the remaining challenges and integrated improvement strategies toward Si-based full cells.