Predicting stress fIelds of porous metamaterials under arbitrary rotations is essential for reliable online monitoring in additive manufacturing. However, existing neural operators are limited by two major issues: geometric orientation bias where identical lattices are misinterpreted under rotation, and non-physical stress leakage caused by the failure of Partial Differential Equation (PDE) constraints at discontinuous void-solid interfaces in porous metamaterials. To address these challenges, we propose the Symmetry-Imbued Hamiltonian Neural Operator (SIHNO) that embeds geometric symmetry and an energy-structured inductive bias into a unified neural operator architecture. Specifically, SIHNO introduces a Rotation-Steerable Lattice Projector (RSLP) that lifts lattice images into a continuous-angle geometric symmetry representation. This geometric representation is further coupled with a Hamiltonian-Fueled Propagator (HFP) that replaces local PDE constraints with Hamiltonian-inspired energy propagation. Finally, a slice-aware convolutional decoder reconstructs stress fields based on RSLP and HFP. Comprehensive experiments demonstrate that SIHNO outperforms existing neural operators. The proposed framework provides a robust architecture for stress-field prediction in additive manufacturing.

Lean manufacturing is now called Industry 4.0, in which traditional production tools are data-driven and have implications for corporate sustainability. Jidoka (JIDO) has been transformed into JIDO 4.0, which employs Internet of Things (IoT) systems and sensors to gather data from the production process for real-time monitoring and decision making. Based on three hypotheses, this study proposes a structural equation model (SEM) to examine the relationships between JIDO 4.0, digital sustainability (DISU), and environmental sustainability (ENSU) in manufacturing companies. The SEM was tested with data from 834 responses to a questionnaire for managers, engineers, and supervisors, validated using Lawshe’s content validity ratio and Aiken’s V. The Warp3 algorithm in WarpPLS 8.0 was used to detect nonlinear relationships between constructs. The results indicate that the three proposed hypotheses are supported, showing that JIDO has the greatest direct effect on DISU (β = 0.588), and DISU is the most influential predictor of ENSU (β = 0.553). The indirect effect of JIDO on ENSU, mediated by DISU (β = 0.325), was nearly equal to the direct effect (β = 0.342), indicating that DISU acts as a bridge between intelligent autonomy and environmental benefits. A sensitivity analysis based on conditional probabilities indicated that when DISU was high, the probability of achieving a high ENSU was 64.8%. These results indicate that JIDO is a data-driven organizational capability with direct and indirect effects on ENSU, and that it provides managers with empirical evidence to prioritize their investments in smart and lean technology.

Predictive maintenance for industrial equipment is critical for improving production safety, reducing maintenance costs, and optimizing equipment utilization. However, existing deep learning methods face two key challenges in industrial equipment prognostics: the lack of uncertainty quantification to support risk-informed decision-making and the inability to simultaneously capture multi-scale temporal patterns in equipment degradation processes. This paper presents the Temporal Probabilistic Joint Embedding Predictive Architecture (TP-JEPA), a novel deep neural network framework that learns robust representations of equipment health states by predicting probabilistic distributions of future states in latent space. TP-JEPA’s innovations include: (1) a probabilistic encoding mechanism that extends deterministic representations to distributions, inherently quantifying prediction uncertainty; (2) a multi-scale temporal encoder designed to extract hierarchical features from high-frequency transients to long-term degradation trends; and (3) a multi-task learning paradigm that jointly optimizes anomaly detection, remaining useful life (RUL) estimation, and health state assessment, enabling synergistic task enhancements. Evaluations on the National Aeronautics and Space Administration (NASA) bearing dataset demonstrate that TP-JEPA achieves an Area Under the Receiver Operating Characteristic curve (AUROC) of 0.9999 for anomaly detection—outperforming state-of-the-art methods—and a mean absolute error of 69.1 cycles for remaining useful life prediction, with well-calibrated uncertainty estimates (95% confidence interval coverage of 94.6%). Cross-dataset validation and ablation studies confirm the framework’s efficacy and robustness.