In Industry 5.0, human-centric smart manufacturing prioritizes the needs of technologists to help enterprises sustain competitive advantages. In this context, design for manufacturing (DFM) plays an essential role, as it ensures the manufacturability of digital designs to deliver high-quality products. Due to novice designers’ limited manufacturing knowledge, the implementation of DFM depends on repeated and passive design iterations, placing a heavy burden on designers. Existing research on improving DFM focuses on manufacturability analysis, which only provides analysis results but ignores novice designers’ manufacturability needs for design modifications. To bridge the gap, this paper proposes a novel human-centric proactive DFM approach that aims to address designers’ manufacturability needs throughout the design process to reduce passive iterations and meet evolving industry demands. Specifically, considering multiple design parameters, a deep learning network is trained for 3D model generation and similarity calculation. Next, the learned network can support human-centric proactive DFM, which includes two parts: automated manufacturability guidance for incomplete designs and manufacturability analysis for complete designs. Through 3D model generation, incomplete designs can be completed and unmanufacturable designs can be modified. Furthermore, similarity calculation facilitates historical manufacturable case recommendation to meet designers’ needs in their decision-making. Experimental results show the efficacy of the approach, achieving accuracy improvements of 4.17% on the impeller dataset and 4% on the manufacturing feature dataset in manufacturability analysis, compared with state-of-the-art approaches. Application examples demonstrate its effectiveness to assist novice designers to proactively improve product manufacturability.

Human-robot collaboration (HRC) assembly necessitates precise mutual cognition to guarantee safe and efficient execution. In this context, human assembly intention recognition (HAIR) serves as a critical approach to achieving this mutual understanding. However, most current HAIR approaches struggle to extract sufficient spatiotemporal information from limited industrial data, particularly under complex conditions like varying scales and visual occlusions. Thereby, this article proposes an ensemble encoder approach to extract and fuse spatial and temporal features from visual and skeleton streams of the HRC assembly process, thus significantly improving HAIR accuracy and efficiency. First, an RGB feature extraction encoder is designed to model spatiotemporal dependencies of the assembly process with different scales of features from flexible input RGB encoders (RGBEs). Distinctively, a cross-attention module is utilized to fuse information from different-scale RGBEs, ensuring comprehensive assembly action representation with different granularities. Second, to address the occlusion challenge, a mask-aware skeleton feature extraction encoder is devised. By utilizing frame and joint masking strategies, it robustly models the relationship between operator pose evolution and assembly actions, maintaining high performance even under occlusion. Third, a global feature fusion encoder integrates and aligns features from RGB and skeleton feature extraction encoders. Experimental results demonstrate the state-of-the-art performance of the proposed approach, which achieves the highest accuracy of 99.12%, 99.23%, and 84.59% on MCV-Intention, HA4M, and HA-VID datasets, respectively. Six ablation studies demonstrate the performance effects of fusion positions, the number of depth channels, cross-attention fusion module, occlusions, illuminations, and computational efficiency.

Proactive human-robot collaborative (HRC) assembly has caught great attention as emerging paradigm for flexible mass personalization in manufacturing with respect to Industry 5.0. To realize adaptive and ergonomic collaboration, it is essential to enable robots recognize human assembly intention based on context-aware information precisely at edge side with Industrial Internet of Thing, also known as human assembly intention recognition (HAIR). For this purpose, this paper systematically reviewed the most relevant papers from major digital databases, where 127 papers are investigated with designed search procedure that published until July 2024. And reviewed papers are summarized from the perspective of: (1) assembly scene perception based on multimodalities data; (2) understanding of HAIR based on machine learning; (3) HAIR application for HRC process. In addition, four current challenges and future research trends are also discussed to facilitate full-adaptive and mutual-cognitive HRC assembly environments.

Design is a complex process and multifaceted knowledge is required in a collaborative environment. Especially, knowledge modeling and application across design and manufacturing domain can reduce automotive development iteration. However, it still a challenge for designer-oriented manufacturing knowledge reuse due to the lack of relationship capture for design contexts and their manufacturing constraints. To solve this problem, the paper presents a rule-enhanced collaborative design method, which can infer manufacturing constraints required in the specific design context. First, a framework of collaborative design is proposed. Then the ontology is used for knowledge modeling and rule construction. Finally, a simple case is illustrated to demonstrate the effectiveness of the method.

Configuring intelligent manufacturing systems (IMSs) is significant for manufacturing enterprises to take a step toward Industry 4.0. However, most current IMS is configured based on the Industrial Internet of Things (IIoT) with a centralized architecture, which results in poor flexibility to handle manufacturing disturbances and limits capacity to support security solutions. To solve the above issues, this article combines IIoT with the permissioned blockchain and proposes a novel manufacturing blockchain of things (MBCoT) architecture for the configuration of a secure, traceable, and decentralized IMS. Then, hardware infrastructures and software-defined components of MBCoT are designed to provide an insight into the industrial implementation of IMS. Furthermore, the consensus-oriented transaction logic of MBCoT is presented based on a crash fault-tolerant protocol, which empowers MBCoT with a strong but resource-efficient encryption mechanism to support the autonomous manufacturing process. Finally, the implementation of an MBCoT prototype system and its application examples justify that the proposed approach is practical and sound. The evaluation experiment demonstrates that MBCoT equips IMS with a secure, traceable, stable, and decentralized operating environment while achieving competitive throughput and latency performance.

In industrial enterprises, effective retrieval and reuse of three-dimensional (3-D) computer-aided design (CAD) models could greatly save time and cost in new product development and manufacturing. Consequently, this article proposes a novel view-based approach for 3-D CAD model retrieval enabled by deep learning. This article constructs a multiview model dataset in industrial domain that collects solid and line views of database models. Since views contain rich information for differentiating these models, the problem of model retrieval is defined as a view recognition problem. Then, the extended deep residual networks (ResNets) are successfully trained to facilitate the model retrieval. With the learned networks, engineers could take a group of views, an engineering drawing, or even a hand-drawn sketch that represents their query intents as input and acquire the relevant 3-D CAD models and embedded knowledge for product lifecycle reuse. The experimental results demonstrate the effectiveness and efficiency of the approach.