Optimizing three-stage hybrid flow shop scheduling: A dynamic programming and hybrid meta-heuristic framework for joint production and preventive maintenance under real-world constraints

Large language models (LLMs) are gaining increasing popularity in software engineering (SE) due to their unprecedented performance across various applications. These models are increasingly being utilized for a range of SE tasks, including requirements engineering and design, code analysis and generation, software maintenance, and quality assurance. As LLMs become more integral to SE, evaluating their effectiveness is crucial for understanding their potential in this field. In recent years, substantial efforts have been made to assess LLM performance in various SE tasks, resulting in the creation of several benchmarks tailored to this purpose. This paper offers a thorough review of 291 benchmarks, addressing three main aspects: what benchmarks are available , how benchmarks are constructed , and the future outlook for these benchmarks . We begin by examining SE tasks such as requirements engineering and design, coding assistant, software testing, AIOps, software maintenance, and quality management. We then analyze the benchmarks and their development processes, highlighting the limitations of existing benchmarks. Additionally, we discuss the successes and failures of LLMs in different software tasks and explore future opportunities and challenges for SE-related benchmarks. We aim to provide a comprehensive overview of benchmark research in SE and offer insights to support the creation of more effective evaluation tools.

Purpose. This study aims to develop and validate a comprehensive approach for analyzing unstructured textual descriptions of defects extracted from actual aviation maintenance data. The goal is to improve both the efficiency and depth of fault analysis by addressing two key tasks: automatic classification of defects into standard categories and identification of latent thematic subgroups within these categories. Methodology. The research is based on a dataset containing maintenance records from nine commercial aircraft over a seven-year period. A multi-stage preprocessing pipeline was developed, including an algorithm for domain-specific abbreviation identification and expert-driven decoding. To solve the multiclass classification task across 30 Chapter–Section (CS) categories, four approaches were compared: CountVectorizer with LinearSVC, TF-IDF and Word2Vec with logistic regression, and fine-tuning of the transformer-based DistilBERT model. For an in-depth analysis of the largest defect category, topic modeling based on Latent Dirichlet Allocation (LDA) was applied, with a quantitative procedure for selecting the optimal number of topics. Findings. The best performance in classification was achieved by the TF-IDF with logistic regression approach, reaching f1-macro = 0.762 and Cohen’s Kappa = 0.809, statistically comparable to CountVectorizer with LinearSVC. Classical methods significantly outperformed neural network models, underscoring their robustness for analyzing short technical texts. Topic modeling successfully decomposed the largest defect category into five interpretable and semantically coherent subgroups. Originality. The novelty of this work lies in developing and testing a formalised method for analysing unstructured aviation maintenance data, implemented as a single integrated process. The study also provides a detailed comparative evaluation of classical and modern NLP models on domain-specific aviation maintenance data. Practical value. The work is practical in nature and contains results which are ready for implementation. A prototype of an automated classifier has been created which is capable of processing the main flow of daily defect reports, reducing the time required for manual processing. An in-depth failure analysis tool has also been developed, which provides a transition from general fault codes to the analysis of specific sub-problems. This contributes to optimizing maintenance programs, enhancing diagnostic procedures, and ultimately improving flight safety.

A hot fix is an unplanned improvement to a specific time-critical issue deployed to a software system in production. While hot fixing is an essential and common activity in software maintenance, it has never been surveyed as a research activity. Thus, such a review is long overdue. In this paper, we conduct a comprehensive literature review of work on hot fixing. We highlight the fields where this topic has been addressed, inconsistencies we identified in the terminology, gaps in the literature, and directions for future work. Our search concluded with 140 articles on the topic between the years 1986 and 2024. The articles found encompass many different research areas such as log analysis, runtime patching (also known as hot patching), and automated repair, as well as various application domains such as security, mobile, and video games. We find that many directions can take hot fix research forward such as unifying existing terminology, establishing a benchmark set of hot fixes, researching costs and frequency of hot fixes, and researching the possibility of end-to-end automation of detection, mitigation, and deployment. We discuss these avenues in detail to inspire the community to systematize hot fixing as a software engineering activity.

Spiraea japonica is widely used for landscaping in China, particularly in urban gardens and parks. During the summer of 2024, leaf blight symptoms were observed on S. japonica in Tongbai County, Henan Province, China (32°22′21″ N, 113°24′10″ E). Field surveys across different areas recorded a disease incidence of 51% (n = 100 plants). Initial symptoms appeared as small, dark brown to nearly black, circular to irregular leaf spots that expanded with tan centers and darker margins; lesions frequently coalesced, producing large, blighted areas and occasional premature defoliation. Diseased leaves (n = 15) were collected and surface sterilized. From each leaf, two pieces were excised from the lesion margin and placed on potato dextrose agar (PDA). Twenty-seven Colletotrichum isolates were recovered (positive isolation rate 90%); three representative isolates from distinct sites (X-21, X-22, X-23) were selected for detailed analyses. On PDA after 5 days of incubation, colonies were fast-growing with aerial mycelium; colonies exhibited a salmon-pink surface with a vinaceous-red reverse. Conidia were hyaline, aseptate, smooth-walled, and cylindrical to fusiform, measuring 9.1–16.6 × 1.6–5.2 μm (n = 100). Appressoria were medium to dark brown, ovoid with entire margins. These features are consistent with members of the Colletotrichum acutatum species complex (Talhinhas et al. 2023). For molecular identification, five loci—the internal transcribed spacer (ITS), dehydrogenase (gapdh), chitin synthase (chs-1), actin (act), and calmodulin (cal)—were amplified for each isolate. All newly generated sequences have been deposited in GenBank under the following accession numbers: PV342493-PV342495 (ITS), PV798316-PV798318 (gapdh), PV575167- PV575169 (chs-1), PV417281, PV417282, PV468789 (act), PV755113, PV660595, PV755114 (cal). A maximum-likelihood phylogeny based on the concatenated dataset placed X-21, X-22, and X-23 within the Colletotrichum fioriniae clade, clearly separated from other congeners. BLAST results showed ≥99–100% identity to reference C. fioriniae strains for all loci. Together, morphology and multilocus phylogeny support the identification of these isolates as C. fioriniae (Zhang et al. 2023). Pathogenicity was assessed by applying 50 μL droplets of a conidial suspension (10⁶ conidia mL⁻¹) onto the adaxial surface of unwounded leaves on five potted S. japonica plants; leaves on five additional plants received 50 μL of sterile water as mock-inoculated controls. Plants were maintained at 28 °C and ~90% relative humidity, lesion consistent with symptoms observed in the field development was evaluated 3 days post-inoculation, whereas control plants remained symptomless. The pathogen was consistently re-isolated from symptomatic tissues and identified as Colletotrichum spp. based on morphology and ITS sequences, thereby fulfilling Koch’s postulates. C. fioriniae causes anthracnose in various hosts in China, including walnut, persimmon, Camellia oleifera, Rhus chinensis (Zhu et al. 2015; Fan et al. 2024; Liu et al. 2025; Xu et al. 2023). Because S. japonica is extensively cultivated for its decorative foliage and profuse flowering, the emergence of anthracnose poses a tangible threat to aesthetic function in public landscapes. Accurate identification of C. fioriniae is thus critical to guide fungicide selection, and support early intervention in landscape maintenance programs.

The Unity Atlantic Rhythm Map is a digital archive prototype that visually unites Afro-diasporic intangible heritage in performances by American, Caribbean and African culture-bearers (mostly emigrés) – and related media from their homelands and the wider diaspora. Watch the 82 second Teaser video (D. C. Molloy 2024b). In the Unity Map, rhythm refers to music and dance interchangeably and as one, as in West African griot cultures (Traore 2020; Comtet 2012; Molloy 2021). As this journal edition considers Africaneity through Caribbean dance, I share my rhythm mapping process. This reveals living African rhythms and historic population flows that make the Caribbean a major hub of Afro-diasporic dance knowledge circulation. Find my browser-based application – the Unity Atlantic RhythmMap – open access online. Or take a 3-minute video tour of the Unity Map in English (D. C. Molloy 2025b) or French (D. C. Molloy 2025a). The Unity Map offers an audiovisual journey through circles of past, present, and future social dance – dances we carry in muscle memory, as observed by Katrina Hazzard-Donald in her book Jookin’ (1990). The Unity Map curates archival, dance community and social media sources as well as official video channels by iconic artists. Any reliable, ethical source that illustrates rhythm real-time dance-drum syncrony, rhythm-coded vocabulary, imagery, formation and steps with black representation and expertise, will be considered for the Map. Methodologies include videodance and multimedia ethnography. A unifying aesthetic thread is rhythm as a ritual of social identity and individual expression. Map data is organised by recognising rhythm as a unit of embodied collective knowledge, encoded in dance, sound and image. Ancestral rhythm-coded ritual and improvisation emerge as a generative epistemology of Africana memory, identity and belonging. Mapped as a dynamic constellation of knowledge, rhythm iterations illustrate diverse black experiences through local imagery. Explore the Unity Atlantic Rhythm Map https://unityatlantic.org/ (best on tablet or computer with wifi and speakers/headphones). Known bugs: iPhone errors in Map display. A subset of rhythm media is available via Nakala, open access research repository also cited in the Unity Map modal window text tabs. Making Unity Map data independently citable, offers an alternative channel in case the Map interface fails on a given device or on a given day. The Map is decolonial in turning towards indigenous knowledge, prioritising rhythm-based codes of identity over nation-state identity. Slavery made Black Atlantic identity culturally stateless. Nation states historically and periodically exclude or marginalise Afrodescendant peoples and our cultures. The Map offers a unifying Black Atlantic identity narrative that honors our historic trauma, our globally influential rhythm heritage and linguistic diversity. While research and data entry continues, the Map already offers a unique resource for comparing ancestral dances of West Africa with Black dances of the Americas. The Unity Atlantic Rhythm Map is an expression of my Afro-Caribbean identity shaped by a social dance practice spanning many codes, countries and collaborations with Afrodescendant culture bearers. The Map centers real-time drum-dance synchrony, Black representation, virtuoso artists and high production values. The Map's functional longevity depends on software maintenance – so new funding and collaborations need to follow the successful Project Manifest co-production. This article reaches out to funders, curators and dance scholars interested in a Map collaboration or simply in sharing data sources.

To maintain genome stability, proliferating cells must enact a program of telomere maintenance. While most tumors maintain telomeres using telomerase, a subset of tumors utilize a DNA-templated process termed alternative lengthening of telomeres or ALT. ALT is associated with mutations in the ATRX/DAXX/H3.3 histone chaperone complex, which is responsible for deposition of histone variant H3.3 at heterochromatic regions of the genome including telomeres. We wished to better understand the role DAXX plays in ALT suppression, and to determine which disease-associated DAXX mutations are unable to suppress ALT. To answer this question, we leveraged the G292 cell line, in which ATRX is wild-type but DAXX has undergone a fusion event. Restoration of wild-type DAXX in G292 localizes ATRX and abrogates ALT. Using this model system, we tested the ability of disease-associated DAXX missense variants to suppress ALT. Missense mutations in the ATRX binding domain, the histone binding domain, and the C-terminal SUMO interaction motif reduce the ability of DAXX to suppress ALT. Unexpectedly, we find that mutations in the DAXX histone binding domain lead to failure of ATRX localization. We conclude that a key function of DAXX in ALT suppression is the localization of ATRX to nuclear foci.

Predicting bug fix times is a key objective for improving software maintenance and supporting planning in open source projects. In this study, we evaluate the effectiveness of different time series forecasting models applied to real-world data from multiple repositories, comparing local (one model per project) and global (a single model trained across multiple projects) approaches. We considered classical models (Naive, Linear Regression, Random Forest) and neural networks (MLP, LSTM, GRU), with global extensions including Random Forest and LSTM with project embeddings. The results highlight that, at the local level, Random Forest achieves lower errors and better classification metrics than deep learning models in several cases. However, global models show greater robustness and generalizability: in particular, the global Random Forest significantly reduces the mean error and maintains high performance in terms of accuracy and F1 score, while the global LSTM captures temporal dependencies and provides additional insights into cross-project dynamics. The explainable AI techniques adopted (permutation importance, saliency maps, and embedding analysis) allow us to interpret the main drivers of forecasts, confirming the role of process variables and temporal characteristics. Overall, the study demonstrates that an integrated approach, combining classical models and deep learning in a global perspective, offers more reliable and interpretable forecasts to support software maintenance.

Electrical distribution systems constitute vital infrastructure in manufacturing operations requiring high reliability levels to ensure production continuity. This study evaluates the performance of a 2000 kVA transformer and 20 kV cubicle panel at PT Surya Toto Indonesia, Tbk through reliability parameter analysis based on preventive maintenance strategies. A quantitative descriptive method was applied by collecting operational data during July-August 2025 covering 1,464 operating hours. Measured parameters included oil temperature, oil level, input/output voltage, load current, along with failure documentation and repair duration. Reliability analysis employed Mean Time Between Failure (MTBF), Mean Time To Repair (MTTR), and Availability calculations following IEEE Std 493-2017 standards. Research findings demonstrate the transformer achieved MTBF of 360 hours, MTTR of 4.5 hours, and availability of 98.8%, while the cubicle panel obtained MTBF of 292.8 hours, MTTR of 5 hours, and availability of 98.3%. Both availability values exceed Indonesian manufacturing industry standards (>95%) and approach world-class manufacturing category (>98%). Scheduled preventive maintenance programs encompassing daily inspections, weekly, monthly, and annual maintenance proved effective in reducing failure frequency. The study recommends implementing IoT-based condition monitoring and SCADA system integration to enhance reliability toward a 99% target.

Building Information Modelling (BIM) has increasingly transformed global architectural and construction practices by enhancing collaboration, design accuracy, and project efficiency. However, BIM adoption remains slow in several developing countries, including Nigeria, where architectural firms play a critical role in driving digital transformation across the wider construction sector. This study investigates the current level of BIM implementation within Nigerian architectural practices and identifies key factors that either enable or constrain its uptake. Survey findings (77 responses; 77% response rate), analysed using SPSS 26.0 and the Relative Importance Index (RII), reveal that although some firms have begun integrating BIM tools, many still rely heavily on traditional 2D CAD (Computer-Aided Design) workflows. Major barriers include high software acquisition and maintenance costs, limited technical expertise, and insufficient organisational readiness. The results highlight the urgent need for government incentives, targeted capacity-building programmes, and industry-wide digital skill development to accelerate BIM diffusion among architectural firms, whose early adoption is essential for sector-wide modernisation. Future research should explore how socio-technical alignment can reshape BIM-enabled workflows to generate measurable value for clients, contractors, and end users. Examining collaborative data environments, information exchange standards, and participatory design practices will be crucial for demonstrating BIM’s long-term return on investment and establishing sustainable digital transformation pathways within Nigeria’s architectural and construction industries.

This research investigates the production decline and cost increase in the X&Y oil and gas fields from 2022 to 2023. Production fell by 34%, while production costs per barrel rose by 79%. To address these issues, a series of optimization processes are proposed. These aim to restructure and enhance the production facilities to reduce current production costs. The optimizations include reducing pressure at the PPP Platform in 2024. Additional steps include shutting down CPP-ORF 14” pipelines and most processes at CPP 2 Platform by 2027 to convert it into an accommodation platform. Further, the plan involves optimizing the release of two FSO units and directing oil production to ORF via a 16” pipeline by 2028. This study also emphasizes the need for Asset Integrity Management to implement optimization scenarios. This measure aims to reduce maintenance costs, ensure plant integrity, and update maintenance programs based on equipment criticality assessments.

Geothermal Power Plant (GPP) use steam from geothermal reservoirs to produce electricity. Power Plants are one of the most dangerous work environments. The dangers that exist in the power generation process can be in the form of Major Accident Hazard (MAH) and have a large impact which can cause Major Accidents. Major Accident prevention can be done by managing Safety Critical Elements (SCE) as a control measure in the form of procedures to prevent or mitigate the risks. Each SCE has a Performance Standard that includes targets and thresholds set to reduce risk to reach ALARP (As Low as Reasonably Practicable) that will be used as a reference in Inspection, Testing and Preventive Maintenance programs to prevent Major Accidents. This research was conducted to identify SCE and establish Performance Standard for GPP PT XYZ with a capacity of 1x55 MW. Based on the identification results, it is known that the total MAH at GPP PT XYZ is 112. Each MAH was analyzed using a bow-tie diagram to obtain a total of 532 SCE or 8.12% of the total 6,549 existing equipment.

Epidemiology and control of human immunodeficiency and hepatitis C virus in Spanish prisons over the last 30 years: Towards the achievement of sustainable development goals.

This study analyzes the effect of the use of B30/B35 biodiesel on the wear of heavy equipment engine components in the Indonesian mining industry. Through a longitudinal experimental design for 12 months, the study was conducted on 24 units of heavy equipment consisting of excavators, bulldozers, and articulated dump trucks in three different mining locations: a coal mine in East Kalimantan, a nickel mine in Southeast Sulawesi, and a gold mine in Papua. The results show that the use of B30/B35 biodiesel consistently reduces the wear rate of components compared to conventional diesel. Nozzle injector wear in the biodiesel group is 18.6% lower, piston ring wear is 15.3% lower, cylinder liner wear is 12.7% lower, and bearing wear is 14.2% lower. SEM and EDS analysis revealed that biodiesel forms a tribochemical layer on metal surfaces that reduces direct contact between surfaces and minimizes wear. The analysis of the lubricant showed lower concentration of metal particles and better lubricant quality parameters in the biodiesel group. The developed predictive model indicates an extension of component life of around 15-20% with the use of biodiesel, potentially providing maintenance cost savings of 12-18% per year. These findings change the perception that the use of biodiesel is solely regulatory compliance, to an operational strategy that provides economic and technical benefits. This study provides a scientific basis for the optimization of preventive maintenance programs for mining heavy equipment that uses biodiesel and supports the sustainability of the implementation of the national mandatory biodiesel policy.

This study addresses the conflicts between contemporary urban development and the conservation of built heritage in Riobamba (Ecuador), where real estate pressure and lack of maintenance generate tensions within the historic urban fabric. The objective was to analyze how these conflicts are expressed in three representative buildings registered in the inventory of the National Institute of Cultural Heritage: the corner house (Reg. 162), of traditional selective typology and accelerated deterioration that places it at risk of collapse; the republican house (Reg. 071), with emergency interventions that alter its authenticity and reflect social and construction-related pressures; and the eclectic mansion (Reg. 423), which preserves its physical integrity but is affected by an environment transformed by modern buildings. The research followed a qualitative methodology based on heritage technical records, field observation and a review of recent literature. The results show divergent trajectories that reveal how heritage listing, when not supported by economic incentives or appropriate use plans, may restrict demolition without enabling effective intervention processes, thereby favoring abandonment, accelerating degradation and limiting urban renewal. It is concluded that overcoming reactive conservation requires integrated management frameworks that articulate compulsory and co-financed maintenance programs, periodic inspections supported by risk matrices, anti-negligence measures, regulation of urban insertions, adaptive rehabilitation projects and community-based maintenance campaigns. The study provides empirical evidence to rethink heritage governance in intermediate Andean cities and proposes future research lines on digital technologies, life-cycle costs and citizen perception.

The turbine exhaust system is a critical component in maintaining the structural integrity and airworthiness of turbofan-powered transport aircraft on CFM56-7B engines, the Federal Aviation Administration (FAA) issued Airworthiness Directive (AD) 2025-04-03, requiring the modification of the engine primary exhaust nozzle to prevent potential separation and structural damage to the fuselage. This study evaluates the implementation of the directive in the Sriwijaya Air fleet under the scope of Boeing Special Attention Requirements Bulletin 737-78-1106 RB Revision 1. The research method combines airworthiness compliance analysis, engineering inspection, and procedural verification using the Accomplishment Instructions defined in EO M-78-003. The modification involved the installation of TOP KIT 002A0123-1 for Off-Wing Rework Kit and 002A0123-2 for On-Wing Installation Kit, the replacement of affected part numbers 314A2610-1, -62, -68, and incorporation of Airworthiness Limitations (AWLs) Nos. 2, 3, and 4 into the operator’s maintenance program. The results confirm that the affected aircraft—registrations PK-CRI, PK-CRA, PK-CRE, and PK-CRH—belong to Group 1 configuration and require mandatory action within 90 days after the AD effective date April 8, 2025, and must be inspection completed before July 31, 2028.

This paper defines a microeconomic framework for understanding systemic failure in cyber security as market failure. In a marketplace with limited supply chain transparency on software quality in general and software maintenance in particular, rational actors — both software vendors and software buyers — will maximise economic returns by minimising software maintenance and security. As technical debt accrues, so does vulnerability and operational risk, as systems become more difficult to update. In this regard, the depreciation of resilience in software infrastructure is similar to the breakdown of physical infrastructure that is chronically undermaintained, but with the added element of adversarial profit. These problems cannot be solved at the computer science level that created them. They can only be solved as a business problem, as transparency requirements (eg software bill of materials [SBOMs]) and automation slash the cost of diligence, enable preferential selection of higher-quality software and continuous enforcement of terms and conditions for active maintenance.

The oil, gas, and petrochemical industries are under intensifying pressure to operate more safely, reduce environmental risks, and extend the life of critical infrastructure. One of the biggest threats to asset integrity—corrosion under insulation (CUI)—continues to drive costly failures, with CUI reportedly responsible for 60% of pipeline incidents worldwide. This article provides a timely perspective on how the industry is evolving its strategies for managing CUI, as well as the vital role that recommended practices play in aligning inspection, monitoring, and maintenance programs. It highlights the release of CUI-RP-101: Issue 1, the first recommended practice dedicated to continuous CUI monitoring systems. This document was developed with input from major operators including Aramco, Dow, INEOS, Chevron, and CorrosionRADAR.

Pulmonary rehabilitation (PR) is an effective intervention for patients with chronic obstructive pulmonary disease (COPD) but impact typically only lasts 6-12 months. This paper presents results of an economic evaluation of a PR maintenance programme (Self-management Programme of Activity, Coping and Education (SPACE)) undertaken within a prospective assessor-blind randomised controlled trial. Adults with COPD who had completed PR within the previous 4 weeks were randomised to SPACE or best usual care. Healthcare use, personal expenditure and societal costs were recorded at baseline, 6 and 12 months. SPACE costs included staff training, materials and delivery of group sessions. Health utility recorded (EQ-5D-5L) with analysis comparing differences in mean values at 6 and 12 months, over baseline utility scores. Observed changes compared with threshold for COPD clinical significance. Incremental cost-effectiveness ratios estimated from National Health Service and societal perspectives. Cost per quality-adjusted life-year (QALY) values compared with willingness-to-pay threshold (≤£30 000). Uncertainties in costs and outcomes incorporated into a sensitivity analysis. Missing values imputed using a Bayesian mixed model with confounders. 116 patients recruited between October 2019 and June 2022 (57 intervention and 59 control). No significant differences at baseline in age, body mass index, smoking, forced expiratory volume in 1 s and health utility (EQ-5D-5L). Mean healthcare costs in the SPACE group were £139.72 lower per patient over 12 months compared with usual care. At 12 months, the SPACE group retained higher (p=0.04) utility value 0.7609 (SE=0.0238) versus control patients 0.6738 (SE=0.0348). The recorded 0.1178 advantage in mean QALY values (p<0.05) is above the threshold (0.051) for COPD significance. Cost-effectiveness acceptability curves indicate a 97% chance of achieving £20 000 per QALY. Patient and societal costs increase this percentage. This study addresses an important gap in current evidence for non-pharmacological COPD interventions. The PR maintenance programme (SPACE) is shown to be highly cost-effective at 12 months. Future research should consider cost-effectiveness of telerehabilitation programmes, as well as tailored digital support beyond 12 months.

Regression testing is an essential component of software maintenance, aimed at ensuring that newly introduced changes do not negatively impact the existing functionality of a system. In today's fast-paced industrial environments, particularly those employing agile methodologies and Continuous Integration/Continuous Deployment (CI/CD) pipelines, the choice of regression testing technique significantly influences project timelines, resource allocation, and product quality. This research investigates and compares five widely adopted regression testing types: Corrective, Retest-All, Selective, Progressive, and Complete Regression Testing. This study's primary objective is to assess each technique's industrial suitability based on key evaluation metrics such as time efficiency, cost of execution, test coverage, automation/tool support, scalability, and risk of fault omission. We adopted a qualitative scoring methodology, grounded in a comprehensive review of several scholarly articles and industry reports. Each testing type was critically analyzed and benchmarked using a comparison matrix to highlight its strengths and limitations. The analysis reveals that while each regression testing method serves distinct use cases, Selective Regression Testing strikes the best balance between efficiency and coverage for modern industrial needs, particularly in projects with frequent releases and constrained testing budgets. The novelty of this study lies in its holistic, literature-backed comparative framework, explicitly tailored to the industry context aspect often overlooked in prior academic evaluations.