Sensor-equipped tool condition monitoring (TCM) is crucial for automated machining, but missing data poses a significant challenge. Existing methods struggle with the complex patterns and substantial data loss common in these dynamic processes. This paper introduces a novel duration-aware part attention mechanism for robust TCM. Unlike existing attention mechanisms, ours explicitly models time-duration dependencies within sensor signals, capturing multi-scale representations of tool degradation even with incomplete data. The part-attention operator, adapted from the Swin Transformer, can dynamically weight different time segments based on their duration and relevance. We further incorporate a cross-dimensional self-attention mechanism to fuse information across multiple sensors and time steps, capturing complex relationships indicative of tool wear. We evaluate our method on real-world machining datasets with varying levels of missing data, demonstrating its superior ability to accurately monitor tool condition compared to existing methods. The results show that the duration-aware part-attention effectively captures crucial temporal dependencies, leading to robust TCM even with substantial data loss.

The advancement of the Industry 5.0 in information technology has led to increased interest in integrating edge-cloud cooperation with Internet of Things (IoT) and cyber-physical system (CPS) designs. This integration effectively reduces service delays and provides real-time analysis feedback to physical spaces, attracting attention from both academia and industry. These advanced technologies enhance production system intelligence, their alignment with circular economy principles for promoting sustainability has been overlooked. To address this gap, this article proposes an intelligent sustainability evaluation strategy driven by edge-cloud cooperation, IoT, and CPS. The proposed approach aims to enhance production sustainability and intelligence through circular economy perspectives. It introduces improved gray relation analysis and deep clustering network techniques to extract meaningful insights from diverse indicators within the evaluation system. By analyzing relationships between different equipment and workshops, it provides an analytical method that enhances production efficiency while reducing energy consumption and resource waste. To further validate the proposed method, an illustrative example using a partner company’s production data demonstrates its accuracy.

To determine the real-time changes in brain arousal introduced by anesthetics, Electroencephalogram (EEG) is often used as an objective neuroimaging evidence to link the neurobehavioral states of patients. However, EEG signals often suffer from a low signal-to-noise ratio due to environmental noise and artifacts, which limits its application for a reliable estimation of depth of anesthesia (DoA), especially under high cross-subject variability. In this study, we propose an end-to-end deep learning based framework, termed as AnesFormer, which contains a data selection model, a self-attention based classification model, and a baseline update mechanism. These three components are integrated in a dynamic and seamless manner to achieve the goal of improving the effectiveness and robustness of DoA estimation in a leave-one-out setting. In the experiment, we apply the proposed framework to an office-based dataset and a hospital-based dataset, and use seven existing models as benchmarks. In addition, we conduct an ablation experiment to show the significance of each component in AnesFormer. Our main results indicate that 1) the proposed framework generally performs better than the existing methods for DoA estimation in terms of effectiveness and robustness; 2) each designed component in AnesFormer is likely to contribute to the DoA classification improvement.

The potential intelligence behind advanced machining systems (AMSs) offers positive contributions toward process improvement. Imitation learning (IL) offers an appealing approach to accessing this intelligence by observing demonstrations from skilled technologists. However, existing IL algorithms that implement single policy strategies have yet to consider realistic scenarios for complex AMS tasks, where the available demonstrations may have come from various experts. Moreover, most IL assumes that the expert's policy is optimal, preventing the learning from fulfilling the previously ignored green missions. This article introduces a novel three-phase policy search algorithm based on IL, enabling the learning of heterogeneous expert policies while balancing energy preferences. The first phase equips the agent with machining basics through upper-level policy learning, generating an imitation policy distribution with various decision-making principles. The second phase enhances energy conservation capabilities by employing Pareto-improvement learning and fine-tuning the agent's policies to a Pareto-policy manifold. The third phase produces outcomes and amplifies the efficacy of human feedback by utilizing ensemble policies. The experimental results indicate that the proposed method outperforms meta-heuristics, exhibiting superior solution quality and faster computation times compared to four diverse baseline methods, each with diverse samples.

With the growing demands on green short life-cycle products, advanced energy-aware process planning (AEPP) becomes critical. A major limitation of the existing methods is the poor resistance to the perturbations encountered in advanced machining systems. Therefore, a graph convolutional reinforcement learning (GCRL) method is proposed to overcome such limitations. In this method, a graph convolutional policy network is trained to rapidly adapt the learned commonalities to specific tasks. Unlike algorithms that fix decision variables before optimization, this method employs graph generation to represent AEPP while taking into consideration the flexibilities of operations, machines, and cutting tools. The problem is reformulated as a novel Markov decision process (MDP) to describe the dynamic generation procedure of process plans. A graph convolutional network (GCN) is concurrently used to perform graph embedding to compress the topology of input graphs. Additionally, reinforcement learning (RL) is used to achieve robust and intuitive learning for process planning. To improve the adaption performance of the proposed GCRL, a two-phase multitask training strategy is adopted. Learning efficiency is improved because agents can incorporate both intertask similarities and task-specific rules. A comprehensive case study, including energy characteristics and algorithm performance analyses, is also performed to validate the developed method.

Energy efficiency modeling is of great importance to energy management and conservation for machinery enterprises. To improve the generalization ability, this article combines the machining parameters and the configuration parameters into energy efficiency models, for which machine-learning (ML) algorithms are used considering the lack of theoretical formulas. Based on the three-year data collected in a shop floor, a comparative study for two different cases is conducted with a particular focus on prediction accuracy, stability, and computational efficiency. In Case 1, only cross-sectional data are used to predict energy efficiency, ignoring the deterioration of spindle motors and cutting tools. Three traditional ML algorithms, i.e., artificial neural networks, support vector regression, and Gaussian process regression, are evaluated with the help of five error metrics. In Case 2, we construct the models in a more realistic situation that considers the dynamic aspects of spindle motor aging and tool wear. A convolutional neural network, a stacked autoencoder, a deep belief network and the aforementioned traditional ML algorithms are investigated. The comparison shows that all the models in Case 1 suffer from performance degradation, while deep learning achieves the long-term improvement in accuracy. Note to Practitioners-Energy efficiency models deliver many advantages, ranging from energy-aware machine design to process optimization. Although a large amount of works in the past focused on physics-based and experimental modeling for specific machining configurations, it can be more effective to improve the applicability of the modeling methods by involving the configuration variables into the models. Due to the uncertainties in both the machine and the operation environment, machine learning is adopted to fit the high-dimensional and high-nonlinear energy system. To the best of our knowledge, this is the first article that provides a comprehensive survey on ML-based modeling in terms of data sizes, temporal granularities, feature selection, and algorithm performance. Such a survey helps engineers quickly justify the appropriate ML methods to meet the actual requirements.

Energy-efficient machining has become imperative for energy conservation, emission reduction, and cost saving of manufacturing sectors. Optimal machining parameter decision is regarded as an effective way to achieve energy efficient turning. For flexible machining, it is of utmost importance to determine the optimal parameters adaptive to various machines, workpieces, and tools. However, very little research has focused on this issue. Hence, this paper undertakes this challenge by integrated meta-reinforcement learning (MRL) of machining parameters to explore the commonalities of optimization models and use the knowledge to respond quickly to new machining tasks. Specifically, the optimization problem is first formulated as a finite Markov decision process (MDP). Then, the continuous parametric optimization is approached with actor-critic (AC) framework. On the basis of the framework, meta-policy training is performed to improve the generalization capacity of the optimizer. The significance of the proposed method is exemplified and elucidated by a case study with a comparative analysis.

This paper studies a predictive modeling for cutting energy consumption in CNC turning process by using deep learning methods. An analysis of energy consumption in cutting period is firstly presented, based on which the impact factors of energy are clarified. Then the data collection platform and data pre-processing are introduced, followed by a brief review of Convolutional Neural Network (CNN), Stacked Auto-Encoder (SAE) and Deep Belief Network (DBN). These modeling methods are tested by k-fold cross-validation. The obtained results show that SAE is the most suitable method to model the relationship between process parameters, machining configuration and cutting energy.

This paper studies the energy characteristics with respect to process parameters from a systematic point of view, in terms of electricity and materials. A detail analysis of energy characteristics of a CNC machining system is firstly presented, based on which the calculation models of energy efficiency are formulated. Then the effects of process parameters on energy and processing time are investigated by using S/N analysis. The results show different optimization trends for two kinds of specific energy consumption considered in this work and detail explanations of the trends are given afterwards.