Programme technique 2025

Voici le résumé des présentations et les biographies des conférenciers. Les présentations peuvent être français ou en anglais. La langue du titre et du résumé indique la langue de chaque présentation.

L’HORAIRE TECHNIQUE EST ACTUELLEMENT EN COURS D’ÉLABORATION ET SERA ULTÉRIEUREMENT DISPONIBLE EN TÉLÉCHARGEMENT

From Vibration to Insight: Advancing Predictive Maintenance with AI and Data-Driven Diagnostics

As industrial systems grow in complexity and interdependence, traditional maintenance strategies fall short in meeting the demands for resilience, cost-efficiency, and real-time decision-making. This plenary presentation explores cutting-edge predictive maintenance strategies that harness deep learning, remaining useful lifetime (RUL) prediction, and optimization to improve reliability and resilience in industrial systems.
We explore how advanced models, such as Bi-LSTM neural networks and probabilistic RUL estimators, transform raw vibration signals into actionable prognostics for rotating machinery. Examples will demonstrate how these models enable early fault detection, optimized maintenance timing, and reduced unplanned downtime. Key insights include: i) integration of vibration data into deep learning architectures for accurate fault prediction; ii) use of uncertainty-aware RUL prediction to guide risk-informed maintenance decisions; iii) scalable optimization models that turn health predictions into cost-effective maintenance schedules for fleets and complex systems.

Claver Diallo, Dalhousie University, Keynote Speaker/Conférencier invité

Claver Diallo, Ph.D., P.Eng., is a Professor in the Department of Industrial Engineering at Dalhousie University in Halifax, Nova Scotia, where he has taught since 2007.  His research focuses on performance optimization in production and service systems, with expertise in maintenance engineering and management, reliability and availability engineering, remanufacturing, project scheduling, and sustainable closed-loop supply chain design.  Dr. Diallo’s recent work explores the integration of reliability engineering and predictive maintenance with smart production and hyperconnected logistics systems within the Industry 4.0/5.0 paradigm.  He holds a Ph.D. and M.A.Sc. in Industrial Engineering, as well as a B.Eng. in Mechanical Engineering, from Laval University (Québec). He is a member of the Institute of Industrial and Systems Engineering (IISE), Engineers Nova Scotia, Canadian Operational Research Society (CORS), and International Federation of Automation and Control (IFAC) Technical Committee 5.2.

Claver Diallo, PhD, ing., est professeur au Département de génie industriel de l’Université Dalhousie à Halifax (Nouvelle-Écosse), où il enseigne depuis 2007. Ses travaux de recherche portent sur l’optimisation de la performance des systèmes de production et de service, avec une expertise en ingénierie et gestion de la maintenance, ingénierie de la fiabilité et de la disponibilité, remanufacturage, ordonnancement de projets et conception de chaînes d’approvisionnement durables. Ses travaux récents explorent l’intégration de l’ingénierie de la fiabilité et de la maintenance prédictive à la production intelligente et aux systèmes logistiques hyperconnectés dans le cadre du paradigme de l’industrie 4.0/5.0. Il est titulaire d’un doctorat, d’une maîtrise et d’un baccalauréat en génie mécanique, de l’Université Laval (Québec). Il est membre de l’Institute of Industrial and Systems Engineering, Engineers Nova Scotia, Société canadienne de recherche opérationnelle (SCRO) et de la Fédération internationale de l’automatisation et du contrôle (IFAC-TC 5.2).

L’avenir de la surveillance de l’état de santé des actifs dans un monde en pénurie de main-d’œuvre

Les organisations investissent depuis des années dans les technologies et la formation afin de mettre en œuvre et de maintenir des programmes de surveillance de la santé des actifs. Cependant, la disponibilité des ressources qualifiées pour soutenir ces activités essentielles diminue. En conséquence, les organisations explorent diverses solutions pour combler cet écart et accomplir davantage avec moins de ressources. Dans ce contexte, nous examinerons différentes possibilités pour l’avenir de la surveillance de la santé des actifs, telles que le suivi à distance, l’intégration de l’intelligence artificielle et de l’apprentissage automatique, la surveillance en ligne comparativement à la collecte manuelle des données, et bien plus encore. Nous évaluerons également le retour sur investissement de ces différentes options.

The future of Condition Monitoring and Reliability in a world with skill shortage

Asset-intensive organizations have been investing in technologies and training for years to implement and maintain robust condition monitoring programs, yielding significant benefits. However, the availability of skilled resources to sustain these critical activities is diminishing. As a result, organizations are exploring various solutions to bridge this gap and accomplish more with fewer resources. In this context, we will explore different possibilities for the future of condition-based monitoring (CBM), such as remote CBM, AI and ML integration, online CBM vs.versus route-based monitoring, and more. We will also assess the return on investment for these options.

Alain Pellegrino, Reliability Solutions, Keynote Speaker/Conférencier invité

Alain Pellegrino est un professionnel certifié en maintenance et fiabilité (CMRP) fort de plus de 20 ans d’expérience comme consultant en fiabilité. Il est président de Reliability Solutions LP, à Pensacola, en Floride. Il dirige et gère une équipe de 50 consultants techniques spécialisés en maintenance de précision, inspections opérateurs et ingénierie de fiabilité. Leur mission est d’aider les clients à atteindre des résultats profitable grâce à l’excellence en matière de fabrication fiable.

Alain Pellegrino is a Certified Maintenance and Reliability Professional (CMRP), with more than 20 years’ experience in Reliability consulting. He is the President of Reliability Solutions LP, in Pensacola Florida. Leading and managing a team of 50 technical consultants, specialized
in precision maintenance, Operator care and reliability engineering. Their mission is to assist clients in achieving Profitable Results® through excellence in Reliable Manufacturing®.

Identifying and Solving Resonance, A Case Study – DPS Pump

This case study follows the entire process, from initial discovery to the final solution, of identifying and solving a resonance problem involving a dry pit submersible pump. The presentation begins with how the problem was presented to us along with the client’s frustration at not having a solution that would work, even though some had been tried. We show how we developed a test plan to identify and confirm the problem using multiple tests including bump testing, runup and coast down testing, and phase analysis using operating deflection shapes. We then present the test results data using a variety of graphics including data plots, photographs, and animations. Relevant details of the AINSI / Hydraulic Institute 9.6.4 specification and the OEM vibration specification are reviewed and discussed. There is also some explanation of the hydraulic forces generated by centrifugal pumps and the vibration that results. Finally, the options available to solve resonance in this circumstance are discussed. Some are not possible, and some are not acceptable to the client. Given these conditions, a solution is determined and implemented. The solution is presented with more photographs of how it was tested and finally implemented. Many of the original tests were reperformed on the solution to substantiate the effectiveness of the solution.

Grant Akitt, PDM Technologies

Grant is a Mechanical Engineer specializing in vibration analysis on rotating equipment and their structures. With over 30 years of experience in the industry, he is accomplished in vibration measurement and analysis, transient data analysis, operational deflection shape testing / phase analysis, and dynamic balancing.
Grant began his career in 1991 as a Condition Monitoring Specialist for SKF Canada Limited in Scarborough, Ontario. He then worked for SKF Condition Monitoring, based in Houston, Texas, for 3 years as an Industry Specialist in both the Metals and Mining group and the Pulp and Paper group. Following this was a variety of positions working as a Team Lead in the Predictive Maintenance field for service companies operating in Southern Ontario, after which he acquired PDM Technologies in 2007. Grant is a licensed Professional Engineer and a certified Category III vibration analyst since 2017.

Grant est un ingénieur en mécanique, spécialisé dans l’analyse des vibrations des équipements rotatifs et de leurs structures. Fort de plus de 30 ans d’expérience dans l’industrie, il est reconnu pour ses compétences en mesure et analyse des vibrations, en analyse de données transitoires, en essais de déformation en fonctionnement. en analyse de phase et en équilibrage dynamique. Grant a débuté sa carrière en 1991 comme spécialiste en surveillance conditionnelle chez SKF Canada Limitée à Scarborough, en Ontario. Il a ensuite travaillé pendant trois ans chez SKF Condition Monitoring, à Houston, au Texas, comme spécialiste industriel au sein des groupes Métaux et mines et Pâtes et papiers. Il a ensuite occupé divers postes de chef d’équipe en maintenance prédictive pour des entreprises de services opérant dans le sud de l’Ontario, avant d’acquérir PDM Technologies en 2007. Grant est un ingénieur agréé et un analyste en vibrations certifié de catégorie III depuis 2017.

Développement d’une méthode de mesure de couple sans contact sur un arbre d’un groupe turbine alternateur d’hydro-électricité

Les augmentations de puissances des centrales existantes prévues par Hydro-Québec dans le plan d’action 2035 nécessitent des études approfondies de faisabilité afin d’assurer la sécurité du personnel et de l’installation. Les rejets de charges sont particulièrement d’intérêt en raison de l’augmentation potentielle de la force des phénomènes transitoires alors que le groupe est déconnecté du réseau. Dans le cadre de ces études, afin de bien caractériser le comportement de la turbine, il est requis de mesurer avec précision le couple en temps réel sur un arbre dont le diamètre peut être de plus de 4 pieds. Les méthodes conventionnelles de mesure de couple se font à l’aide de jauges de contraintes et nécessitent des systèmes de télémétrie ou encore des contacts tournants. Ce type d’installation nécessite invariablement des équipements spécialisés et plusieurs heures de travail. Dans la recherche d’alternative à ce type d’instrumentation, une nouvelle méthode de mesure de couple sans contact basée sur la mesure visuelle de l’angle de torsion a été testée à l’aide d’un montage dédié puis les apprentissages ont été transposés pour une utilisation réelle.

Mathieu Soares, Hydro-Québec

Mathieu Soares est un technicien expert en mécanique au Centre de recherche d’Hydro-Québec (CRHQ), spécialisé dans le domaine vibro-acoustique depuis 2001. Diplômé en Technologie Physique, il a développé une expertise approfondie en participant à de nombreux projets de recherche et développement, ce qui lui a permis d’atteindre un niveau élevé de connaissances et d’expérience en vibro-acoustique. Il a collaboré avec des chercheurs et des technologues de renommée mondiale dans ce domaine. Au fil des années, il a étudié divers types d’équipements liés à l’hydroélectricité, au transport et à la distribution d’électricité. Grâce à son travail au CRHQ, il a contribué au développement de nouvelles méthodes de pronostic des équipements d’Hydro-Québec. Il est membre de l’ACVM et analyste catégorie II.

Mathieu Soares is an expert mechanical technician at the Hydro-Québec Research Center (CRHQ), specializing in the field of vibro-acoustics since 2001. With a degree in Physics Technology, he has developed in-depth expertise by participating in numerous research and development projects, which has allowed him to achieve a high level of knowledge and experience in vibro-acoustics. He has collaborated with world-renowned researchers and technologists in this field. Over the years, he has studied various types of equipment related to hydroelectricity, as well as the transmission and distribution of electricity. Through his work at the CRHQ, he has contributed to the development of new prognostic methods for Hydro-Québec’s equipment. He is a member of the ACVM and a category II analyst.

Optimising resource-constrained fleet selective maintenance with asynchronous maintenance breaks

This research offers a novel and significant extension of the Fleet Selective Maintenance Problem (FSMP) by considering asynchronous maintenance breaks and resource-constrained maintenance planning. This constitutes a significant shift from the conventional focus on synchronous breaks for the FSMP formulated to plan maintenance for fleets of mission-critical systems. This paper establishes a theoretical link between the Selective Maintenance Problem (SMP) and the Resource Constrained Project Scheduling Problem (RCPSP). The proposed FSMP formulation for asynchronous breaks is more general and versatile in adapting to a broad spectrum of operational constraints and resource scarcities. Numerical experiments are conducted that highlights the trade-offs between the timing and quality levels of maintenance activities and the consumption of resources that maintenance planners can make to obtain the best system performance for the budget and maintenance windows available.

Alireza Amjadian, University of Dalhousie

Alireza Amjadian is currently pursuing a Ph.D. degree from the Department of Industrial Engineering at Dalhousie University, Halifax, Nova Scotia, Canada. He holds a Master of Applied Science degree and a Bachelor’s degree in Industrial Engineering from Kharazmi University, Iran. His current research interests include remanufacturing, pricing and revenue management, supply chain, and intelligent maintenance operations.

Alireza Amjadian poursuit actuellement un doctorat au département de génie industriel de l’Université Dalhousie, à Halifax, Nouvelle-Écosse, Canada. Il est titulaire d’une maîtrise ès sciences appliquées et d’un baccalauréat en génie industriel de l’Université Kharazmi, en Iran. Ses intérêts de recherche actuels portent sur le remanufacturage, la tarification et la gestion des revenus, la chaîne d’approvisionnement et les opérations de maintenance intelligente.

Using Bi-LSTMs for Diagnostics and Prognostics in Data-Driven Maintenance Planning

Modern industries dependent on reliable asset operation under constrained resources employ intelligent maintenance methods to maximize efficiency. However, classical maintenance methods rely on assumed lifetime distributions and suffer from estimation errors and computational complexity. The advent of Industry 4.0 has increased the use of sensors for monitoring systems, while deep learning (DL) models have allowed for accurate system health predictions, enabling data-driven maintenance planning. Often, sensor data comes in a time-series format, making recurrent neural networks (RNNs) particularly suitable for data analysis. A subset of RNNs that have shown a lot of promise in diagnostic and prognostic tasks is bidirectional long short-term memory models (Bi-LSTMs). However, past literature using Bi-LSTMs in these tasks has not fully tested the architecture’s limits, either by limiting the scope of data used or by failing to explicitly tune hyperparameters, instead augmenting with additional DL architecture without justification. Hence, this paper tests a minimalist Bi-LSTM model with the Case Western Reserve University dataset to assess its ability to perform multi-faceted bearing fault diagnoses. The model has seven hyperparameters explicitly tuned prior to testing and has performance on par with many state-of-the-art models. With the strength of Bi-LSTM architecture demonstrated, a hybrid Bi-LSTM model is developed to generate RUL predictions from the NASA C-MAPSS dataset and a separate filter dataset. Using Monte Carlo dropout, empirical reliability functions are generated for the optimization of the selective maintenance problem (SMP). The proposed framework is used to plan maintenance for a mission-oriented series k-out-of-n:G system. Numerical experiments compare the framework’s performance against prior SMP methods and highlight its strengths. When minimizing cost, maintenance plans frequently result in mission survival while avoiding unnecessary repairs. The proposed method is usable in large-scale, complex scenarios and finds exact solutions while avoiding the need for computationally-intensive parametric reliability functions.

Alexandros Noussis

Alexandros Noussis is a graduate student studying industrial engineering at Dalhousie University, based in Halifax, Nova Scotia. He gained his Bachelor of Industrial Engineering in 2023 and Master of Applied Science in 2025, both at Dalhousie. His MASc researched focused on condition-based, intelligent maintenance for marine renewable energy (MRE) systems while incorporating machine learning. He is currently pursuing his PhD on intelligent maintenance planning for MRE production assets. His work in MRE has expanded to include data collection frameworks and digital twin use. He is a recipient of the 2023 Natural Sciences and Engineering Research Council of Canada (NSERC) Canada Graduate Scholarship for Master’s students, as well as a recipient of the 2025 Postgraduate Scholarship for Doctoral students from NSERC.

Alexandros Noussis est un étudiant diplômé en génie industriel de l’Université Dalhousie, située à Halifax, en Nouvelle-Écosse. Il a obtenu son baccalauréat en génie industriel en 2023 et sa maîtrise en sciences appliquées en 2025 de Dalhousie. Ses recherches de maîtrise en sciences appliquées portaient sur la maintenance intelligente et conditionnelle des systèmes d’énergie marine renouvelable (EMR) en intégrant l’apprentissage automatique. Il poursuit actuellement un doctorat sur la maintenance intelligente des systèmes de production d’énergie renouvelable. Ses travaux en EMR se sont élargis pour inclure la collecte de données massives et l’utilisation des jumeaux numériques. Il est lauréat de la bourse d’études supérieures du Canada pour les étudiants à la maîtrise du Conseil de recherches en sciences naturelles et en génie du Canada (CRSNG) en 2023, ainsi que de la bourse d’études supérieures pour les étudiants au doctorat du CRSNG en 2025.

Optimizing budget allocation for multimission selective maintenance planning

Mission-critical systems in sectors such as aerospace, defence, transportation, petrochemistry, and power generation require high reliability to prevent failures causing major economic losses, environmental damages, and safety risks. For such systems, solving the selective maintenance problem (SMP) yields optimal maintenance planning decisions during scheduled breaks. Its extension, the multi-mission SMP (MMSMP),focuses on optimizing component maintenance, maintenance levels, and repairperson assignments over multiple consecutive missions interspersed with maintenance breaks. While recent advances integrate predictive,resource-constrained, and fleet-wide strategies, they rely on the unrealistic assumption of fixed budgets, ignoring the reality of fluctuating and tight financial constraints faced by planners. This study investigates how different maintenance budget allocations across missions affect system performance. Using a two-phase de-composition model and binary integer programming, it explores various budget distribution strategies: uniform,linearly increasing, and inverted-V. The goal is to determine how allocating resources differently across mis-sions can enhance asset reliability within fixed budget limits. The findings aim to guide maintenance plannersin making budget decisions to improve overall system reliability while balancing resource constraints.

Farzad Falahaty, University of Dalhousie

Farzad Falahaty is currently pursuing a MASc degree from the Department of Industrial Engineering at Dalhousie University, Halifax, Nova Scotia, Canada. He holds a Master of Management of Technology degree from Iran University of Science and Technology (IUST) and a Bachelor’s degree in Industrial Engineering from IAU South Tehran Branch Faculty of Engineering, Iran. His current research interests include Maintenance Engineering and Management, Reliability and Availability Engineering, and Quality Control & Reliability.

Farzad Falahaty prépare actuellement une maîtrise en sciences au département de génie industriel de l’Université Dalhousie, à Halifax, en Nouvelle-Écosse, au Canada. Il est titulaire d’une maîtrise en gestion des technologies de l’Université iranienne des sciences et technologies (IUST) et d’une licence en génie industriel de la faculté de génie de l’antenne sud de Téhéran de l’IAU, en Iran. Ses recherches portent actuellement sur l’ingénierie et la gestion de la maintenance, l’ingénierie de la fiabilité et de la disponibilité, ainsi que le contrôle qualité et la fiabilité.