Risk-Based Maintenance Optimization

The food processing industry is undergoing a profound transformation with the integration of Artificial Intelligence (AI), evolving from traditional automation to intelligent, adaptive systems aligned with Industry 5.0 principles. This review examines AI’s role across the food value chain, including supply chain management, quality control, process optimization in key unit operations, and emerging areas. Recent advancements in machine learning (ML), computer vision, and predictive analytics have significantly improved detection in food processing, achieving accuracy exceeding 98%. These technologies have also contributed to energy savings of 15–20% and reduced waste through real-time process optimization and predictive maintenance. The integration of blockchain and Internet of Things (IoT) technologies further strengthens traceability and sustainability across the supply chain, while generative AI accelerates the development of novel food products. Despite these benefits, several challenges persist, including substantial implementation costs, heterogeneous data sources, ethical considerations related to workforce displacement, and the opaque, “black box” nature of many AI models. Moreover, the effectiveness of AI solutions remains context-dependent; some studies report only marginal improvements in dynamic or data-poor environments. Looking ahead, the sector is expected to embrace autonomous manufacturing, edge computing, and bio-computing, with projections indicating that the AI market in food processing could approach $90 billion by 2030.

The proliferation of unmanned aerial vehicle (UAV) swarms presents critical challenges to system-level reliability and safety. Traditional maintenance strategies, designed for single assets, are fundamentally inadequate for the systemic complexities and multifaceted risks of swarm operations. This study addresses this gap by developing a multi-objective optimization framework to derive optimal maintenance policies for heterogeneous UAV swarms. We formulate the problem to simultaneously minimize maintenance cost, maximize mission reliability, and minimize a composite operational risk encompassing both crash and hazardous material release. The framework distinguishes between nodal and non-nodal UAV roles and employs the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to explore the complex trade-off space. The framework is validated through numerical experiments against a heuristic benchmark, yielding significant results. Optimized policies reduce injury risk by more than tenfold compared to traditional methods, while simultaneously doubling mission reliability. Furthermore, the analysis reveals a distinct “efficiency frontier” for safety investment, providing a novel, data-driven tool for managerial decision-making. Ultimately, this research delivers a holistic, risk-informed framework that bridges the gap between theoretical optimization and the practical challenges of safe, reliable, and cost-effective swarm operation, offering actionable guidance for tailoring maintenance strategies to specific risk tolerance levels.

Integrating probabilistic risk assessment with the partially observable Markov decision process to consider risk constraints in maintenance optimization

An optimal joint maintenance and mission abort policy for a system executing multi-attempt missions

Joint Optimization of Maintenance and Spare Parts Management in Upstream–Downstream Systems under Quality Control

Joint optimization of maintenance and resource preparation of system with multi-indicator degradation based on multiple failure mode division

Efficacy and safety of ketamine maintenance therapy in treatment-resistant depression: A systematic review of treatment protocols and clinical outcomes.

Reliability-based operation and maintenance optimization for offshore wind farms: Modeling and application in Taiwan

Addressing climate change necessitates innovative and data-driven approaches for monitoring, managing, and reducing greenhouse gas (GHG) emissions across various sectors. Digital Twin Technology (DTT)—which creates dynamic, real-time digital replicas of physical systems— emerges as a transformative tool in this context. This work explores the potential of DTT to accelerate emission reduction efforts in key domains including energy, manufacturing, transportation, and urban infrastructure. By enabling real-time monitoring, predictive maintenance, and system-level optimization, DTT enhances operational efficiency while minimizing environmental impact. The study presents a comprehensive review of current literature, proposes a machine learning–based methodology for forecasting emissions, and introduces a robust evaluation framework. Case studies from recent implementations are analyzed to demonstrate the tangible benefits and scalability of DTT in achieving decarbonization goals. Findings confirm that Digital Twin Technology plays a critical role in advancing toward net-zero emissions and fostering long-term sustainability.

PT. Bumi Khatulistiwa Bauksit (PT. BKB), a bauxite mining company, faces challenges in maintaining its tailing ponds. The lack of a structured maintenance schedule has led to sediment accumulation, which risks pond shallowing and environmental contamination from potential overflow. This study aims to evaluate the condition of the settling ponds and establish an optimal maintenance schedule. A quantitative approach was employed, combining primary field data with secondary data. Analytical steps included calculating design rainfall, rainfall intensity, total discharge, solid percentage, settling velocity, and maintenance/dredging time based on pond volume and sediment accumulation. Results indicate the highest daily total discharge in Pond 1 (17,276.23 m³/day) and the lowest in Pond 5 (5,022.97 m³ /day). Maintenance intervals vary: Pond 1 requires dredging every 1 month and 7 days; Pond 2, every 1 month and 22 days; Pond 3, every 1 year and 5 months and 12 days; Pond 4, every 4 months and 18 days; and Pond 5, every 6 years and 5 months and 12 days. The excavator’s productivity reaches 447.93 m³/hour, with dredging time per pond ranging from 8 to 35 days.

For the design, implementation, performance optimization, and predictive maintenance of high-speed real-time control systems with sub-millisecond control periods, the capability to acquire large volumes of high-rate control data in real time is required without interfering with normal control operation that is repeatedly executed in each extremely short control cycle. In this study, we propose a control-data acquisition method for high-speed real-time control systems with sub-millisecond control periods, in which control data are transferred to an external host device via Ethernet in real time. To enable the transmission of high-rate control data without disturbing the real-time control operation, a multicore microcontroller unit (MCU) is adopted, where the control task and the data transmission task are executed on separately assigned central processing unit (CPU) cores. Furthermore, by applying a double-buffering algorithm, continuous Ethernet communication without intermediate waiting time is achieved, resulting in a substantial improvement in transmission throughput. Using a control card based on TI’s multicore MCU TMS320F28388D, which consists of dual digital signal processor cores and one connectivity manager (CM) core, the proposed control-data acquisition method is implemented and an actual experimental environment is constructed. Experimental results show that the double-buffering transmission achieves a maximum throughput of 94.2 Mbps on a 100 Mbps Fast Ethernet link, providing a 38.5% improvement over the single-buffering case and verifying the high performance and efficiency of the proposed data acquisition method.

Purpose The article aims to address the challenge of partial or complete absence of maintenance data records for industrial assets by generating synthetic maintenance data under a high-quality maintenance data structure established in the framework of International Organization for Standardization (ISO) 14224:2016. The preceding contributes to maintenance engineering, a strategy to obtain meaningful synthetic data in maintenance management analysis without exposing industrial assets to failures that may lead to undesired consequences. Design/methodology/approach The research was conducted under an experimental study aimed at generating synthetic maintenance data from historical statistical distributions of industrial assets. For experimental purposes, based on the criticality of the studied process context, the research was carried out on a centrifugal pump, with its primary data source from the Offshore Reliability Data Handbook (OREDA), from which the four failure modes with the highest failure rate and the non-maintainable components related to the failure rate by probability were selected. The data were processed using Python 3.10.12, using a methodology of standardizing the data structure, for which a pseudo-code was established. Findings The article addresses the generation of synthetic maintenance data using historical statistical distributions from the OREDA. Two sets of synthetic data were obtained for a centrifugal pump, with the second set maintaining originality by defining the maximum failure rate as the mean of the global failure rate based on accurate data, demonstrated with an error of 1.96%. This approach allows for objective decision-making when forecasting different scenarios, as the synthetic data set acquires its dynamics dependent on the statistical distribution of the failure rate by failure modes, evidenced by the error in the standard deviation. Originality/value The article focuses on generating synthetic maintenance data by developing an algorithm based on internationally recognized statistical distributions aligned with the international standards of ISO 14224:2016. This approach aims to create a synthetic maintenance dataset with maintenance records from which maintenance variables and indicators can be derived. These derived insights enable maintenance optimization through data-driven decision-making feedback loops.

The application of Total Productive Maintenance (TPM) using the Overall Equipment Effectiveness (OEE) method at PT. DEF, The issue of maintenance has become crucial as it affects not only the reliability of the machines but also the speed and quality of production. This study will explore how the implementation of TPM using the OEE approach can improve machine effectiveness in a production environment, with a focus on timely and optimal maintenance strategies.The research aimed to assess the efficiency and effectiveness of machine operations, particularly injection molding machines, within the company's production line. Through the collection and processing of operational data, it was found that the OEE value for the machines in question ranged from 94% to 95%, surpassing the company's set standard of 93%. The high OEE score reflects effective machine utilization, availability, performance, and quality rates. The research highlights that while the current system meets international standards for machine maintenance, continuous improvements can further optimize machine productivity and reduce downtime. Keywords: Injection, Molding, OEE, TPM

ABSTRACT In hot strip rolling, even moderate variations in transfer bar (TB) thickness can critically affect the final strip profile and surface quality. This study investigates the influence of roll wear in the last roughing stand (R5) on these inconsistencies by integrating plant measurements with finite element method (FEM) simulation. Work roll profiles were captured at different stages of a rolling campaign using a Pessometer. These profiles were precisely reconstructed using CAD software and incorporated into the FEM based DEFORM-3D simulation environment to reflect worn roll geometries within a thermo-mechanical framework. A comprehensive FEM model was developed to emulate real mill conditions; from slab discharge at the reheating furnace to deformation through the roughing stands and up to the third finishing stand, where the strip's geometric integrity, especially profile is predominantly established. Simulations conducted with varying degrees of R5 roll wear revealed that roll surface degradation induces non-uniform thickness distribution in the TB. These inconsistencies persist into the finishing mill and lead to shape-related defects such as ridge formation. This integrated approach not only helps to visualise the genesis of shape defect but also provides actionable insights into optimal roll maintenance schedules and control interventions to ensure more uniform strip production.

Railway operators have the double problem of aged infrastructure, which requires cost-effective renovation, and growing risk factors caused by climate change, which increase the risk of failure. Although condition monitoring and predictive, risk-adjusted maintenance are advocated, few organizations, on average, successfully convert data variety into cohesive, decision-supporting risk-prioritization of different assets. In this work, an integrated, advanced framework will be offered to present a risk-adjusted Asset Health Index (AHI) using continuous risk analysis based on WSN, data analytics for RUL, and enterprise-wide integrated risk-aware management planning. This will use literature-reviewed best practices in aggregate risk calculation for Asset Health Index and railway infrastructure health analysis, with an additional "context layer" tailored to assess health based on vulnerability to climate-related risks, including heat buckling, flood and scour, and debris from windstorms. The work will also present decision-support bibliographic libraries for risk practices, aggregation, treatment, and risk thresholds. Two simulation scenarios will also be included to show how climate risk accelerates health degradation and changes prioritization. This research provides a approach blueprint checklist to help railway organizations leap over data fragmentation and move forward in smart, risk-based optimization of infrastructure maintenance.

Ravulizumab drug monitoring has not been explored for maintenance therapy in patients with complement-mediated (atypical) hemolytic uremic syndrome (aHUS). Phase III trials suggest the standard dosing regimen provides troughs about threefold higher than needed to suppress complement activity. We describe the use of ravulizumab in pediatric patients with aHUS in remission, exploring potential modified dosing strategies based on serum drug levels and complement markers. This single-center, retrospective cohort study included pediatric patients with aHUS in remission receiving outpatient ravulizumab infusions between June 30, 2023, and March 31, 2024, with at least one ravulizumab trough. Patients received a standard (SR) or a modified (MR) regimen, determined by the nephrologist. The primary outcome was to describe troughs and corresponding AH50 for patients on at least two equal doses. Secondary outcomes included comparison of troughs by regimen, intra-patient variability, possible adverse drug events (pADE), and drug costs. Nine patients were included. The mean ravulizumab trough level was 399.1 (± 107.3) mcg/mL. All patients exceeded the goal of 175 mcg/mL and achieved AH50 < 10%. Four patients (44%) received ravulizumab according to a MR. No difference was observed in ravulizumab trough levels between SR and MR groups (P = 0.67). Patients with multiple troughs showed low intra-patient variability (CV < 25%). pADE rates were similar across regimens, and MR was associated with lower drug costs. Individualized maintenance regimens of ravulizumab based on trough and complement monitoring appear safe and effective while reducing drug costs. Further study is needed to define the optimal ravulizumab maintenance dosing strategy.

Deep Learning (DL), the Internet of Things (IoT), Machine Learning (ML), and Artificial Intelligence (AI) are all combining to change maritime engineering. This study offers a condensed application architecture with an emphasis on fuel optimization, autonomous navigation, and predictive maintenance. Results using simulated datasets and machine learning techniques show a 12% increase in energy efficiency and 94% prediction accuracy in identifying mechanical defects. Predictive models’ quantitative results are displayed in tables and graphs. The study comes to the conclusion that intelligent technology adoption greatly improves maritime systems’ sustainability, safety, and mechanical dependability.

Optimization of maintenance and spares inventory with dependence between the system and environment

An optimal predictive inspection and maintenance policy for a multi-state system: A belief-based SMDP approach