Jae-Chul Lee

This study presents an LSTM autoencoder-based vibration anomaly detection system developed using normal operating data for the servomotor driving the ball screw in a CNC machine tool. The system comprises wireless vibration data acquisition and storage, preprocessing, time series feature generation, model training, and identification of potential anomalies based on reconstruction error (RE). Input features, model architectures, and loss configurations were compared using normal operating data from CNC-A. Vibration RMS and Moving Average were selected as the final input features, together with a loss function incorporating an additional RMS reconstruction error term. The existing standardization parameters and trained models were applied without further adjustment to 29,001 test sequences generated from normal operating data from CNC-B, which had not been used for training or model selection. Across three random seeds, the coefficient of determination (R²) for Vibration RMS was 0.938569 ± 0.000682 across all time steps and 0.931685 ± 0.005062 at the final time step. Potential anomalies were identified using the 99 th percentile (P99) and 99.5 th percentile (P99.5) of the RE distribution derived from the CNC-A normal validation data. The proportion of CNC-B test sequences exceeding P99 ranged from 2.121% to 2.262%, and 567 sequences (1.955%) were identified by all three models. This study demonstrates the implementation and application of a system that transfers models trained on normal operating data and validation-derived thresholds to a different CNC machine tool to quantify reconstruction discrepancies and identify sequences requiring further examination.

A Standard Time Unit (STU), defined as standard work time per unit workload, supports production planning and man-hour estimation in shipbuilding. This study developed a Combined Standard Time Model (CSTM) that reconstructs outfitting work standards by recombining reusable work actions. Process analysis structured outfitting work into five Basic Standard Units (BSUs), 16 Minimum Work Units (MWUs), and 54 Factors (six quantity-type and 48 condition-type). The same fixed-order MWU candidate set was applied to all ten evaluated materials, while material-specific procedures were represented through MWU selection and Factor settings. Base MWU time units and Factor values were derived from existing STU records supplemented by expert judgment. The absolute percentage errors for the ten materials ranged from 0.178% to 2.514%, and all were below 3%, with a mean of 1.281% and a standard deviation of 0.839%. The paired t-test yielded t = 1.617 (p = 0.140), the Pearson correlation was r = 0.999, and Bland–Altman analysis showed a mean difference of 0.305 min with estimated 95% limits of agreement from −0.863 to 1.473 min. Applied without re-adjustment to 13 further materials not used in parameter setting, the model gave a mean absolute percentage error of 3.387% against the existing unit-rate values. Thus, the CSTM maintained numerical consistency with the existing STUs while enabling a standardized labor-time framework and component-level traceability across the evaluated materials. Because model parameterization and evaluation used the same STU records, the results demonstrate internal reconstruction rather than independent predictive superiority; validation against measured work times remains necessary.

Dynamic events, such as new task arrivals, UAV failures, and changes in task time windows, require timely updates to task allocation plans in multi-UAV systems. Global replanning provides broad flexibility for feasibility recovery, but it may also modify assignments and route structures that remain valid after an event. This study proposes Assignment-Preserving Replanning (APR), which inherits the operational state at a common event time, partitions unfinished tasks into directly affected and initially unaffected sets, and explicitly protects valid assignments and relative route structure during repair. When feasibility cannot be recovered within this protected scope, APR selectively releases protected tasks associated with constraint conflicts to expand the modifiable scope. APR was evaluated against constructive repair, partial replanning with a fixed or localized scope, and two global adaptive large neighborhood search methods. Among 116 primary scenarios feasible for all five methods, APR achieved the lowest reassignment rate of 17.71%. It preserved 100.00% of initially unaffected task assignments in 103 scenarios and 100.00% of comparable induced route structure in 91 scenarios. These structural gains were accompanied by a higher mission cost. Across the broader primary evaluation set, APR showed lower feasibility than global replanning but substantially shorter computation time. A stochastic robustness analysis using 10 independent random seeds reproduced the overall trends. Overall, APR provides a complementary replanning strategy that trades some search flexibility and mission efficiency for stronger structural continuity of valid portions of the existing plan. Whether these gains reduce energy consumption, control effort, communication burden, operator workload, or execution risk remains to be validated.

This study develops a time-domain Rankine source method for predicting wave loads and motion responses of adjacent floating bodies. The method combines desingularized Rankine source nodes on the free surface with an acceleration potential formulation to improve the stability of time-domain simulations. The boundary-value problem is formulated using an indirect boundary integral equation method, and the hydrodynamic forces are coupled with the equations of motion. The developed code is validated through convergence tests and comparisons with the experimental data, 3D HOBEM results used as comparison solutions, and HydroStar results for two adjacent barges. The computed wave-exciting forces and motion RAOs reproduce the overall frequency-dependent trends and dominant response characteristics of the reference results. The simulations show that gap flow and shielding effects strongly govern the hydrodynamic interaction between adjacent bodies. In beam seas, the weather-side barge generally experiences larger loads and motions, whereas the lee-side barge is shielded in the short-wave region. Under oblique waves, three-dimensional wave interaction produces a more complex load distribution. Parametric studies show that the separation distance changes the frequency range of gap-flow-induced responses, while shallow water shifts the main resonance toward lower frequencies and modifies the response magnitude. These results demonstrate the applicability of the present method to closely spaced floating systems.

This study presents a real-time framework for generating two-dimensional (2D) orthomosaic maps directly from UAV video. The method targets operational scenarios in which a continuously updated 2D overview is required during flight or immediately after landing, without relying on time-consuming offline photogrammetry workflows such as structure-from-motion (SfM) and multi-view stereo (MVS). The proposed procedure incrementally registers sparsely sampled video frames on standard CPU hardware using classical feature-based image registration. Each selected frame is converted to grayscale and processed under a fixed keypoint budget to maintain predictable runtime. Tentative correspondences are obtained through descriptor matching with ratio-test filtering, and outliers are removed using random sample consensus (RANSAC) to ensure geometric consistency. Inter-frame motion is modeled by a planar homography, enabling the mapping process to jointly account for rotation, scale variation, skew, and translation that commonly occur in UAV video due to yaw maneuvers, mild altitude variation, and platform motion. Sequential homographies are accumulated to warp incoming frames into a global mosaic canvas, which is updated incrementally using lightweight blending suitable for real-time visualization. Experimental results on three UAV video sequences with different durations, flight patterns, and scene targets report representative orthomosaic-style outputs and per-step CPU runtime statistics (mean, 95th percentile, and maximum), illustrating typical operating behavior under the tested settings. The framework produces visually coherent orthomosaic-style maps in real time for approximately planar scenes with sufficient overlap and texture, while clarifying practical failure modes under weak texture, motion blur, and strong parallax. Limitations include potential drift over long sequences and the absence of ground-truth references for absolute registration-error evaluation.

A three-dimensional frequency-domain ship-motion solver based on the Rankine source method is extended to predict added resistance in waves. Although middle-field formulations have been used mainly in time-domain Rankine panel methods, a middle-field evaluation is implemented here within a frequency-domain Rankine source framework and its validity is examined, including low-speed conditions where the enforcement of radiation conditions is challenging. To enhance robustness at low forward speeds, a hybrid radiation technique is incorporated. Convergence studies are carried out for the free-surface and radiation-boundary discretization, as well as for the control-surface resolution and the clearance distance, and practical numerical settings for added-resistance computations are established. The approach is first verified for Wigley III hulls by comparing motion RAOs and added resistance with published experimental and numerical results. It is then validated for the blunt KVLCC2 hull at the design speed and at low speeds (0 and 4 knots) against published measurements and calculations. Further validations are conducted for additional hull forms (Wigley I, KCS, S-175, and Series 60). The results indicate that the proposed frequency-domain Rankine source method with middle-field evaluation and hybrid radiation yields consistent predictions of motion responses and added resistance over a range of speeds and hull forms, while retaining computational efficiency.

Unlike conventional bearings, magnetic bearings (MBs) provide non-contact support and high reliability, but standardized evaluation criteria for marine environments have not yet been established. This study proposes and validates a standardized evaluation procedure for MB-based rotating machinery intended for marine applications. The proposed procedure consists of six steps: (1) equipment data preparation, (2) standardized simplified model generation, (3) magnetic force incorporation, (4) external load definition based on the sea state, (5) analysis (transient dynamic analysis under the defined external loads), and (6) result evaluation. A simplified turbo compressor model was constructed using ANSYS Mechanical, applying equivalent stiffness and damping coefficients to represent magnetic bearing dynamics. External loads were derived from World Meteorological Organization (WMO) Sea State and DNV-RP-C205 guidelines and applied as impulse-type accelerations. Numerical simulation results showed that rotor displacement remained within 95% of the air gap under normal conditions (Sea state 6, 0.8g, here, g = 9.8 m/s2). Under extreme loading (Sea state 8, 1.8g), the analysis indicated that the air-gap limit was exceeded, leading to contact with the backup bearings and casing. The proposed standardized simplified evaluation procedure provides a practical framework for assessing MB-based marine rotating machinery, supporting rapid design verification and classification approval.

We propose a phase-aware complex-spectrogram autoencoder (AE) for preprocessing raw vibration signals of rotating electrical machines. The AE reconstructs normal components and separates fault components as residuals, guided by an input-phasor phase-orthogonality regularization that defines parallel/orthogonal residuals with respect to the local signal phase. We use a U-Net-based AE with a mask-bias head to refine local magnitude and phase. Decisions are based on residual features—magnitude/shape, frequency distribution, and projections onto the normal manifold. Using the AI Hub open dataset from field ventilation motors, we evaluate eight representative motor cases (2.2–5.5 kW: misalignment, unbalance, bearing fault, belt looseness). The preprocessing yielded clear residual patterns (low-frequency floor rise, resonance-band peaks, harmonic-neighbor spikes), and achieved an area under the receiver operating characteristic curve (ROC-AUC) = 0.998–1.000 across eight cases, with strong leave-one-file-out generalization and good calibration (expected calibration error (ECE) ≤ 0.023). The results indicate that learning to remove normal structure while enforcing phase consistency provides an unsupervised front-end that enhances fault evidence while preserving interpretability on field data.

An unsupervised anomaly-detection framework is proposed and field validated for thrust-bearing monitoring in the impulse turbine of a shoreline oscillating water-column (OWC) wave energy converter (WEC) off Jeju Island, Korea. Operational monitoring is constrained by nonstationary sea states, scarce fault labels, and low-rate supervisory logging at 20 Hz. To address these conditions, a 24 h period of normal operation was median-filtered to suppress outliers, and six physically motivated time-domain features were computed from triaxial vibration at 10 s intervals: absolute mean; standard deviation (STD); root mean square (RMS); skewness; shape factor (SF); and crest factor (CF, peak divided by RMS). A feature-based autoencoder was trained to reconstruct the feature vectors, and reconstruction error was evaluated with an adaptive threshold derived from the moving mean and moving standard deviation to accommodate baseline drift. Performance was assessed on a 2 h test segment that includes a 40 min simulated fault window created by doubling the triaxial vibration amplitudes prior to preprocessing and feature extraction. The detector achieved accuracy of 0.99, precision of 1.00, recall of 0.98, and F1 score of 0.99, with no false positives and five false negatives. These results indicate dependable detection at low sampling rates with modest computational cost. The chosen feature set provides physical interpretability under the 20 Hz constraint, and denoising stabilizes indicators against marine transients, supporting applicability in operational settings. Limitations associated with simulated faults are acknowledged. Future work will incorporate long-term field observations with verified fault progressions, cross-site validation, and integration with digital-twin-enabled maintenance.