Rapid advances in new generation information technologies, such as big data analytics, internet of things (IoT), edge computing and artificial intelligence, have nowadays driven traditional manufacturing all the way to intelligent manufacturing. Intelligent manufacturing is characterised by autonomy and self-optimisation, which proposes new demands such as learning and cognitive capacities for manufacturing cell, known as the minimum implementation unit for intelligent manufacturing. Consequently, this paper proposes a general framework for knowledge-driven digital twin manufacturing cell (KDTMC) towards intelligent manufacturing, which could support autonomous manufacturing by an intelligent perceiving, simulating, understanding, predicting, optimising and controlling strategy. Three key enabling technologies including digital twin model, dynamic knowledge bases and knowledge-based intelligent skills for supporting the above strategy are analysed, which equip KDTMC with the capacities of self-thinking, self-decision-making, self-execution and self-improving. The implementing methods of KDTMC are also introduced by a thus constructed test bed. Three application examples about intelligent process planning, intelligent production scheduling and production process analysis and dynamic regulation demonstrate the feasibility of KDTMC, which provides a practical insight into the intelligent manufacturing paradigm.

Pre-existing knowledge buried in manufacturing enterprises can be reused to help decision-makers develop good judgements to make decisions about the problems in new product development, which in turn speeds up and improves the quality of product innovation. This paper presents a graph-based approach to knowledge reuse for supporting knowledge-driven decision-making in new product development. The paper first illustrates the iterative process of knowledge-driven decision-making in new product development. Then, a novel framework is proposed to facilitate this process, where knowledge maps and knowledge navigation are involved. Here, OWL ontologies are employed to construct knowledge maps, which appropriately capture and organise knowledge resources generated at various stages of product lifecycle; the Personalised PageRank algorithm is used to perform knowledge navigation, which finds the most relevant knowledge in knowledge maps for a given problem in new product development. Finally, the feasibility and effectiveness of the proposed approach are demonstrated through a case study and two performance evaluation experiments.

Resistance spot welding poses potential challenges for automotive manufacturing enterprises with regard to ensuring the real-time and accurate quality detection of each welding spot. Nowadays, many machine learning and deep learning methods have been proposed to utilize monitored sensor data to solve these challenges. However, poor detection results or process interpretations are still unaddressed key issues. To bridge the gap, this paper takes the automotive bodies as objects, and proposes a resistance spot welding quality online detection method with dynamic current and resistance data based on a combined convolutional neural network (CNN), long short-term memory network (LSTM), and an attention mechanism. First, an overall online detection framework using an edge–cloud collaboration was proposed. Second, an online quality detection model was established. In it, the combined CNN and LSTM network were used to extract local detail features and temporal correlation features of the data. The attention mechanism was introduced to improve the interpretability of the model. Moreover, the imbalanced data problem was also solved with a multiclass imbalance algorithm and weighted cross-entropy loss function. Finally, an experimental verification and analysis were conducted. The results show that the quality detection accuracy was 98.5%. The proposed method has good detection performance and real-time detection abilities for the in-site welding processes of automobile bodies.