SIC 2025

An appropriate maintenance policy is essential for reducing costs and risks related to equipment failures. A key aspect when defining such policies is the ability to predict the reliability of the systems under study based on a well-fitted model. When developing methods to determine the optimal interval for preventive-maintenance interventions, it is necessary to adopt the model that best represents the reality of the data analyzed and to estimate parameters for the repair effects of preventive and corrective maintenance. This makes it possible to calculate reliability indicators to predict future failure behavior and associated process costs, determine the optimal maintenance interval and ensure the required performance. A real dataset involving failures of mining equipment used by a Brazilian mining company is analyzed, providing valuable information for decision-making.

This study presents the results of the 2024 Priority Route project, focused on improving the reliability and availability of an excavator fleet. Through detailed analysis of the machines’ operating and maintenance conditions, critical points directly affecting productivity were identified, enabling the development of effective strategies to mitigate failures and increase operational efficiency. The Priority Route model, based on simulation, preventive maintenance and optimized resource planning, produced significant results, including increases in MTBF (Mean Time Between Failures) and physical availability, along with a substantial reduction in downtime and number of failures. These advances reflect a considerable improvement in operational performance, positively affecting both cost reduction and excavator reliability.

Human Reliability does not refer to the person alone, but to the reliability of the activity performed, considering the three factors that influence it. Therefore, analyses capable of generating qualitative and quantitative information about the tasks addressed by Human Reliability Analysis (HRA) are required. From this analysis, it is possible to identify tasks with unrecoverable errors that may result in impacts and to guide actions for reducing associated risks, preventing fatalities and catastrophic accidents. This paper presents the application of HTA (Hierarchical Task Analysis), SHERPA (Systematic Human Error Reduction and Prediction Approach) and PSF (Performance Shaping Factors).

This paper presents two complementary approaches for sizing strategic spare-parts inventories, focusing on reliability and the application of advanced data-analysis techniques. The first part describes a reliability study applied to critical pumping systems using Recurrent Data Analysis (RDA) and Reliability, Availability and Maintainability (RAM) analysis, supported by Weibull++ and BlockSim. The second part presents a scalable Machine Learning methodology developed in the R language and patented to forecast consumption and size inventories for thousands of strategic spare parts. Both approaches seek to optimize operational availability and reduce inventory-related costs, contributing to efficient management of industrial assets.

This project aims to implement a Strategic Maintenance Plan for belt conveyors based on reliability studies, increasing asset availability and process maturity. The initiative arose from shortcomings in materials management, maintenance history, inspections and planning, as well as opportunities identified in the database used to monitor shutdowns. The adoption of deliverables based on reliability analysis seeks to reduce corrective interventions, improve safety and optimize operational efficiency, ensuring more sustainable and reliable management.

The reliability of diesel engines in sugarcane harvesters is essential to ensure high operational availability and reduce maintenance costs. The results demonstrate that predictive strategies, such as fluid analysis and monitoring of the failure rate (λ), make it possible to anticipate failures and optimize maintenance planning. The study highlights that adopting reliability-based methodologies can significantly reduce downtime and corrective costs, ensuring greater operational efficiency.

The paper presents a RAM analysis of the seawater intake system of an oil and gas production platform, making it possible to estimate water losses, by volume, resulting from unavailability of its main equipment. These losses affect the cooling system for process equipment and the reservoir water-injection system, causing production losses and therefore reducing project revenue. A case study was conducted using production curves and water-injection quota data to compare different system-design configurations, the performance of three established market technologies for submersible seawater-intake pumps, and aspects related to maintenance strategy. Life Data Analysis (LDA) of the equipment was performed using Weibull++ (2022), based on operational data from production platforms, and the RAM analysis was developed in the event-diagram module of BlockSim (2022).

This study applies quantitative reliability engineering to support strategic decisions in the stabilization plan for ArcelorMittal Tubarão’s sinter cooler. The main system assemblies were analyzed, failures were classified, and indicators such as Mean Time Between Failures (MTBF) and maintenance costs were evaluated. Using methodologies such as Life Data Analysis (LDA), System Reliability Analysis (SRA) and Reliability Growth Analysis (RGA), reliability parameters were modeled in a Block Diagram and tested through Monte Carlo simulation. The approach made it possible to forecast system behavior in subsequent years, quantifying future failures, unavailability and financial impacts. The results support decision-making based on Life Cycle Cost (LCC), providing greater predictability, resource optimization and more efficient allocation of investments in the cooling system.

This article presents the advances and results achieved over the previous two years in Reliability Engineering within Asset Management at Eldorado Brasil, one of the country’s leading pulp companies. The company has demonstrated outstanding performance, repeatedly exceeding its operational records. In 2024, for example, production reached a level 19% above the originally projected volume, directly reflecting growing maturity in industrial asset management. The article describes the current Reliability Engineering structure and its interaction with the other disciplines within the reliability organization. Finally, it analyzes two specific methodologies that, although unconventional, proved highly effective in problem-solving. The objective is to share lessons learned and practices that can be adapted by other organizations.

This paper presents a reliability-based approach to managing the useful life of high-pressure grinding rolls (HPGR), which are essential to the iron-ore pelletizing process. The rolls, manufactured from tungsten carbide, are subject to severe wear, with an impact of up to 15% on production when no spare parts are available; lead time is 365 days. Since 2022, the maintenance strategy has included periodic ultrasonic thickness measurements, enabling predictive analysis. The study uses reliability curves and Life Data Analysis (LDA) to define the maximum wear limit and predict the ideal replacement time. In addition to reducing operational risks, this approach allows more accurate financial planning by integrating maintenance decisions with cash-flow forecasting. The proposed methodology optimizes roll management, minimizes production impacts and reinforces the importance of reliability in the organization’s strategic decision-making.

Study on the application of reliability techniques to develop a roller screen for processing highly cohesive, high-moisture iron ore in a Vale S/A operation in Minas Gerais. The work includes probabilistic analysis, RCA (Root Cause Analysis), RAM Analysis, Reliability-Centered Maintenance (RCM — comparing the conventional method with the use of AI — Artificial Intelligence through Orion) and initial control application. Although this type of equipment is widely used in biomass, scrap and soil industries, its application in iron-ore screening is unprecedented. After design improvements supported by reliability analysis, the screen demonstrated higher productivity than a conventional vibrating screen, with an average gain of 42%. The star screen performed better with hydrated material, efficiently processing cohesive material with moisture above 15%.

Reliability tools play a crucial role in Operational and Maintenance Readiness, ensuring that systems, equipment and processes are prepared to operate efficiently and without failures from the outset. Applying these tools helps identify and mitigate potential risks, ensuring a smooth transition into regular operations. The use of quantitative tools creates demand for a continuous-improvement flow toward the other qualitative tools used during Maintenance Readiness implementation. Integrating all these tools into the Operational Readiness process provides a clear understanding of system risks and reliability and helps managers make informed decisions regarding resource allocation, maintenance planning and risk-mitigation strategies.

RAM (Reliability, Availability, Maintainability) analysis is a methodology used to evaluate the reliability, availability and maintainability of systems and equipment such as the Komatsu 830E2 truck. Reliability refers to the truck’s ability to perform its functions without failure during a specified period. Availability is the probability that the truck will be operational when needed. Maintainability is the ease and speed with which the truck can be maintained or repaired. RAM analysis is essential for identifying and mitigating potential failures, improving operational efficiency and reducing maintenance costs. It helps forecast truck performance and enables informed decisions on preventive and corrective maintenance. In addition, RAM analysis supports the selection of critical components and definition of maintenance strategies, ensuring that the Komatsu 830E meets expected performance and availability requirements.

The importance of reliability has increased exponentially in recent decades. In particular, interest in Human Reliability has grown because of its impact on the reliability of equipment, systems and processes in general. Traditional approaches to reliability modeling may not be sufficient for highly complex systems, especially when little data is available. In such cases, Machine Learning techniques have been used successfully. In this context, the objective is to develop a multivariable model using a hybrid probabilistic-neural approach for Human Reliability analysis.

Coming soon.

The work consists of a cost analysis of the full ballast-cleaning machine used in the ballast-cleaning process on the Estrada de Ferro Carajás. The study considers future demand for use of this asset and evaluates maintenance costs under three different scenarios: acquisition of a new asset, refurbishment of the current asset, or continued use of the same asset. The LCCA methodology was used, including analysis of asset costs under the defined scenarios. In addition, a HAZOP analysis was conducted to define the risks associated with each decision and the corresponding mitigation actions for the selected scenario.

Difficulties in reducing maintenance costs and increasing asset availability are common challenges in maintenance engineering. RAM (Reliability, Availability and Maintainability) analysis is an important methodology for companies with complex systems because it supports informed decisions on maintenance strategies. The analysis optimizes resources and investments and reduces downtime and costs associated with failures. This paper proposes applying RAM Analysis to evaluate and optimize the performance of a crushing system in a mining industrial plant. The methodology includes collection and analysis of historical failure and maintenance data, application of life-data analysis tools, goodness-of-fit tests, block diagrams and RAM analysis itself. The objective is to identify the most critical assets and implement actions to improve operational efficiency, increase system availability and reduce downtime, providing more efficient management of critical assets.

The RCM study process using the Compass AI algorithm will be presented, including the initial study of the equipment’s operating condition and systems, taxonomy review, preparation of the preliminary RCM and validation with specialists. For equipment already in operation, with maintenance tasks already loaded and used in the CMMS, a mapping sheet was developed to prepare the upload sheet and maintenance plans with the lowest possible impact on maintenance execution.

Seven iron-ore mines operate in Vale’s Southern Corridor: Tamanduá, Horizonte, Abóboras, Mina do Pico, Capão Xavier, Fábrica and Viga. Four of them require rock-blasting processes that depend on the reliability of drills responsible for creating holes for explosives. The complex has 13 drills: nine Caterpillar MD6290 units and four Atlas Copco D65 units. Between January and August 2023, the drills maintained physical availability according to plan. However, between September 2023 and March 2024, availability of the MD6290 units fluctuated, affecting mine blasting activities. Analyses identified long-duration corrective interventions related to rotating components as the main problems, with some components reaching only 68% of the useful life stated by the manufacturer. One initiative was to conduct Life Data Analysis (LDA) of the components to define the best replacement strategy. RAM Simulation helped assess impacts on costs, failures and spare-parts stock. Between April and August 2024, the strategy was revised, generating significant improvements in physical availability and MTBF, reducing MTTR and increasing monthly blasted mass by 18.7%.

The project seeks to identify failure modes and their potential impacts on asset reliability and, through the methodology, suggest the result that could potentially be achieved with the implemented improvements.

Brazil’s energy mix is currently composed mainly of hydroelectric plants, which represented an average of 60% of total generation in 2024, requiring high asset reliability to ensure power generation. Using appropriate maintenance strategies is essential, and RAM Analysis is a methodology capable of providing the necessary robustness to these strategies, delivering valuable results for its three pillars: Reliability, Availability and Maintainability. This study focused on performing RAM analysis of three generating units in a hydroelectric power plant using BlockSim and Weibull++ on the systems that made up the fifth level of the hierarchy of those units. The results made it possible to increase plant availability through review of maintenance plans and equipment and, through studies, ensure increased reliability of systems with a high failure-criticality index.

This project applied a quantitative, structured and field-oriented approach to increase the reliability of iron-ore Reclaimer RC-01, reducing critical and recurring failures, controlling operational risks and promoting sustainable performance improvements.

Reliability studies can be important tools in project implementation because they guide the identification of previously unknown problems and, consequently, engineering solutions before startup. To increase decision-making accuracy and ensure operation of a mineral-beneficiation plant fully dependent on the operability of an emergency tailings-containment sump, a system-reliability study was developed for this sump. RAM analysis was performed using data from a similar plant to assess potential bad actors and create a preventive action plan aimed at reducing ramp-up impacts. To ensure the best maintenance strategy for the equipment, RCM was also developed for the most important assets using Artificial Intelligence (AI) to reduce study time. Finally, minimum stock levels for the most important components were evaluated using Spare Parts Forecast (SPF). The estimated avoided loss from the study is approximately R$584,530.22, considering component expenses and avoided production losses, with approximately 223,166.09 tons of avoided production loss.

This paper addresses the application of reliability tools to optimize a drilling-platform maintenance shutdown, focusing on reducing dry-dock time and the total process cost. Through Accelerated Life Testing, degradation analysis and Monte Carlo simulation, together with predictive techniques and fitness-for-service methods, it was possible to identify wear and degradation points and implement wall-thickness monitoring of the asset, ensuring the platform’s operational safety. The results demonstrate that combining reliability tools and predictive techniques can significantly optimize shutdown scope, generating time and cost savings while ensuring operational safety and reliability.

In an environment increasingly driven by data and the pursuit of operational efficiency, reliability becomes an essential competitive differentiator in asset management. This presentation examines how technology can be applied strategically to transform asset management, promoting innovation, predictability and data-driven decision-making. Concepts and tools will be presented that enable integration of engineering, maintenance, operations and information technology, with a focus on solutions based on analysis and understanding of organizational context. The approach is aimed at professionals in technical and managerial leadership roles who seek to enhance results through modern and sustainable asset-management practices. Join the session and discover how to align technology, strategy and performance to achieve excellence in asset management.