Diesel engines remain the foundation for obtaining mechanical energy in sectors where autonomy and reliability are required; however, predictive diagnostics under real-world conditions remain challenging. The purpose of this scoping review is the investigation and systematization of published scientific data on the application of vibration methods for monitoring the technical condition of diesel engines in industrial or controlled laboratory conditions. Based on numerous results of publication analysis, sensor configurations, diagnosed components, signal analysis methods, and their application for assessing engine technical condition are considered. As methods for determining vibration parameters, time-domain and frequency-domain analysis, adaptive decompositions, and machine and deep learning algorithms predominate; high accuracy is more often achieved under controlled conditions, while confirmations of robustness on industrial installations are still insufficient. Key limitations for the application of vibration monitoring methods include the multicomponent and non-stationary nature of signals, a high level of noise, requirements for sensor placement, communication channel limitations, and the need for on-site processing; meanwhile, the assessment of torsional vibrations remains technically challenging. It is concluded that field validations of vibroacoustic data, the use of multimodal sensor platforms, noise-immune algorithms, and model adaptation to the specific environment are necessary, taking into account fuel quality, transient conditions, and climatic factors.

Centrifugal compressors (CCs) are key components of marine power plants (MPPs), supporting engine boosting, boil-off gas (BOG) handling on liquefied natural gas (LNG) carriers, and auxiliary services such as heating, ventilation, and air conditioning (HVAC). However, recent publications are often fragmented by domain (aerodynamics, mechanical design, standards, and digitalization), complicating cross-domain engineering decisions for marine duty cycles. This structured review follows an explicit protocol to synthesize peer-reviewed studies (2015–2025) retrieved from Scopus and Web of Science and organizes the evidence by application class: turbocharger-integrated stages for marine diesel and gas-turbine engines, LNG/BOG compression trains, and auxiliary onboard services. The synthesis consolidates (i) aerodynamic KPIs (pressure ratio, efficiency, surge and stall margins, and operating range), (ii) mechanical and lifecycle enablers (seals, bearings, and rotordynamics), and (iii) quantified impacts of digital methods (control, diagnostics, and digital twins). Reported trends include single-stage pressure ratios of ~5.4–5.7, multistage overall pressure ratios exceeding 10, and surge-margin improvements of ~40–44% associated with advanced diffusers as well as casing and endwall treatments. Industrial case studies (non-marine) report downtime reductions of ~25–35% and maintenance-cost reductions of ~25%, while evaluated diagnostic datasets show high accuracy. Key gaps remain in marine-specific validation datasets and harmonized testing and data standards.

This paper proposes a formal framework for synthesizing multi-stage condition-based maintenance (CBM) decision logic for marine diesel monitoring systems. The design object is treated not as a single threshold or classifier output, but as an implementable decision logic with explicit stages of data-quality gating, thresholding, confirmation, fusion, and temporal filtering. Decision quality is evaluated using unconditional control-reliability indicators (CRIs) under a prescribed prior probability of rare abnormal events within a unified Monte Carlo verification protocol. Within a simplified Gaussian surrogate model, we compare baseline thresholding, repeated-measurement averaging, within-path confirmation, and measurement-level fusion. For the reported reference configuration, averaging five repeated measurements yields the largest reduction in the raw error criterion, “2 out of 3” confirmation provides a smaller but consistent improvement, and two-path multi-fidelity fusion is beneficial only after calibration toward the more informative path. The results show that, under rare abnormal events and limited measurement accuracy, decision quality is determined primarily by calibration of the multi-stage channel-level logic rather than by thresholding alone.

A forced fault label does not establish that diagnostic evidence supports an engineering conclusion. This study evaluates a post-prediction D0–D5 layer combining calibrated class probabilities, predicted severity, pressure–torsion feature-group consistency, and evidence completeness. The simulated 3500-DEFault benchmark at 2500 rpm and 0, 15, 30, and 60 dB was partitioned by scenario into training, calibration, validation, and test sets. Operating parameters were selected on validation data and locked before final-test evaluation. Across five split seeds, the calibrated classifier had error 0.0462 ± 0.0032. The status layer achieved coverage 0.9791 ± 0.0015 and four-class selective risk 0.0342 ± 0.0039. The status false-safe rate was 0.00208 ± 0.00165, compared with the forced false-safe rate of 0.0094 ± 0.0038. Among true G4 cases, 0.9795 ± 0.0024 received D3/D4. Approximately matched entropy rejection had slightly lower selective risk (0.0329 ± 0.0037); superiority in rejection efficiency is not claimed. Partial torsional-feature loss exposed substantial severe-case underwarning without universal refusal. The contribution is an interpretable separation of normal, intermediate, high-confidence fault, severe-fault, and refusal indications. Repeated splits quantify split sensitivity, not independent engine replication. Results are limited to the simulated fixed-speed benchmark and do not independently validate shipboard maintenance decisions.