New Post-Doc Open Position

Title of Research

Artificial intelligence methods for reliability prediction and predictive maintenance

Description of the Research

Predicting the reliability and performing predictive maintenance is fundamental for enhancing the safety and profitability of complex industrial systems. The objective of the research program is to develop methods for reliability prediction and predictive maintenance. Specifically, the research will focus on the development of adaptive and explainable AI models, transfer learning methods and physics informed neural networks. One key issue for the application of AI is the estimation of the uncertainty of the predictions. To this aim, Bayesian Neural Networks, Monte Carlo dropout and other techniques will be also considered. The research will be done within the activities of PE1 FAIR Spoke 4: ADAPTIVE AI – Future Artificial Intelligence Research – CUP D53C22002380006


Prof. Piero Baraldi and Prof. Enrico Zio




The research will be carried out within international collaborative projects with industries, research centers and universities. The research team will be formed by the supervisors, the research fellow and PhD and master thesis students. Periodic discussion meetings will be held to update the supervisors about the work carried out and the corresponding results, and to establish the next activities. The development of the program will take place at the laboratory Lasar ( at the Department of Energy, Bovisa campus.



May 22nd, 2023


Online Services of Politecnico di Milano – section Open competitions and selections – Open competition/selection for allocation of job/post – Calls for temporary research fellowships (see the attached step-by-step guidance)


Euro 28500, gross of fees charged to the contractor (It is expected to amount to approximately 2100 euros per month, after taxes)


2 years


Congratulations to Dr. Zhaojun Hao!

Dr. Zhaojun Hao, now an ex-member of LASAR, has successfully defended his PhD thesis entitled “Operation and Maintenance of Cyber-Physical Energy Systems Accounting for Reliable and Safe Power Production and Supply,” under the guidance of Prof. Zio and Prof. Di Maio.

Dr. Hao will soon return to China, his home country, to explore new opportunities. The LASAR team wishes him the best in his professional achievements!


New Publication!

New Publication!

Title: Optimization of the Operation and Maintenance of renewable energy systems by Deep Reinforcement Learning

Journal: Renewable Energy

Authors: Luca Pinciroli, Piero Baraldi, Guido Ballabio, Michele Compare, Enrico Zio

Citation: Pinciroli, L.; Baraldi, P.; Ballabio, G.; Compare, M.; Zio, E. Optimization of the Operation and Maintenance of renewable energy systems by Deep Reinforcement Learning. Renewable Energy 2022, 183, 752-763.

This is an open-access article that can be downloaded from the following link before Jan 14, 2022:


Equipment of renewable energy systems are being supported by Prognostics & Health Management (PHM) capabilities to estimate their current health state and predict their Remaining Useful Life (RUL). The PHM health state estimates and RUL predictions can be used for the optimization of the systems Operation and Maintenance (O&M). This is an ambitious and challenging task, which requires to consider many factors, including the availability of maintenance crews, the variability of energy demand and production, the influence of the operating conditions on equipment performance and degradation and the long time horizons of renewable energy systems usage. We develop a novel formulation of the O&M optimization as a sequential decision problem and we resort to Deep Reinforcement Learning (DRL) to solve it. The proposed solution approach combines proximal policy optimization, imitation learning, for pre-training the learning agent, and a model of the environment which describes the renewable energy system behavior. The solution approach is tested by its application to a wind farm O&M problem. The optimal solution found is shown to outperform those provided by other DRL algorithms. Also, the approach does not require to select a-priori a maintenance strategy, but, rather, it discovers the best performing policy by itself.


New Publication!

Title: Metamodeling and On-Line Clustering for Loss-of-Flow Accident Precursors Identification in a Superconducting Magnet Cryogenic Cooling Circuit

Journal: Energies


Vincenzo Destino, Nicola Pedroni, Roberto Bonifetto, Francesco Di Maio, Laura Savoldi and Enrico Zio

This is an open-access article that can be downloaded from the following link:


In the International Thermonuclear Experimental Reactor, plasma is magnetically confined with Superconductive Magnets (SMs) that must be maintained at the cryogenic temperature of 4.5 K by one or more Superconducting Magnet Cryogenic Cooling Circuits (SMCCC). To guarantee cooling, Loss-of-Flow Accidents (LOFAs) in the SMCCC are to be avoided. In this work, we develop a three-step methodology for the prompt detection of LOFA precursors (i.e., those combinations of component failures causing a LOFA). First, we randomly generate accident scenarios by Monte Carlo sampling of the failures of typical SMCCC components and simulate the corresponding transient system response by a deterministic thermal-hydraulic code. In this phase, we also employ quick-running Proper Orthogonal Decomposition (POD)-based Kriging metamodels, adaptively trained to reproduce the output of the long-running code, to decrease the computational time. Second, we group the generated scenarios by a Spectral Clustering (SC) employing the Fuzzy C-Means (FCM), in order to identify the main patterns of system evolution towards abnormal states (e.g., a LOFA). Third, we develop an On-line Supervised Spectral Clustering (OSSC) technique to associate time-varying parameters measured during plant functioning to one of the prototypical groups obtained, which may highlight the related LOFA precursors (in terms of SMCCC components failures). We apply the proposed technique to the simplified model of a cryogenic cooling circuit of a single module of the ITER Central Solenoid Magnet (CSM). The framework developed promptly detects 95% of LOFA events and around 80% of the related precursors.


New Publication!

Title: A Framework based on Finite Mixture Models and Adaptive Kriging for Characterizing Non-Smooth and Multimodal Failure Regions in a Nuclear Passive Safety System

Journal: Reliability Engineering & System Safety

Authors: L. Puppo, N. Pedroni, Francesco di Maio, A. Bersano, C. Bertani, Enrico Zio

This is an open-access article that can be downloaded from the following link before October 17:


In the safety analyses of passive systems for nuclear energy applications, computationally demanding models can be substituted by fast-running surrogate models coupled with adaptive sampling techniques; for speeding up the exploration of the components and system state-space and the characterization of the conditions leading to failure (i.e., the system Critical failure Regions, CRs). However, in some cases of non-smoothness and multimodality of the state-space, the existing approaches do not suffice. In this paper, we propose a novel methodological framework, based on Finite Mixture Models (FMMs) and Adaptive Kriging (AK-MCS) for CRs characterization in case of non-smoothness and/or multimodality of the output. The framework contains three main steps: 1) dimensionality reduction through FMMs to tackle the output non-smoothness and multimodality, while focusing on its clusters defining the system failure; 2) adaptive training (AK-MCS) of the metamodel on the reduced space to mimic the time-demanding model and, finally, 3) use of the trained metamodel provide the output for new input combinations and retrieve information about the CRs.

The framework is applied to the case study of a generic Passive Safety System (PSS) for Decay Heat Removal (DHR) designed for advanced Nuclear Power Plants (NPPs). The PSS operation is modelled through a time-demanding Thermal-Hydraulic (T-H) model and the pressure selected for characterizing the PSS response to accidental conditions shows a strong non-smooth and multimodal behavior. A comparison with an alternative approach of literature relying on the use of Support Vector Classifier (SVC) to cluster the output domain is presented to support the framework as a valid approach in challenging CRs characterization.

Title: Accounting for Safety Barriers Degradation in the Risk Assessment of Oil and Gas Systems by Multistate Bayesian Networks

Journal: Reliability Engineering & System Safety


This is an open-access article that can be downloaded from the following link before September 22:


In this paper, a multistate Bayesian Network (BN) is proposed to model and evaluate the functional performance of safety barriers in Oil and Gas plants. The nodes of the BN represent the safety barriers Health States (HSs) and the corresponding conditional Failure Probability (FP) values are assigned. HSs are assessed on the basis of specific Key Performance Indicators (KPIs) related to the barrier characteristics (i.e., technical, procedural or organizational, continuously monitored or event-based characterized). FP values are estimated from failure datasets (for technical barriers), evaluated by Human Reliability Analysis (HRA) (for operational and organizational barriers) and assigned by expert elicitation (for barriers lacking data or information). For illustration, the multistate BN model is developed for preventive barriers and applied to a case study related to the potential release of flammable material in the slug catcher of a representative O&G Upstream plant which may lead to major accident scenarios (fire, explosion, toxic dispersion). The results from the case study demonstrate that the multistate BN model is able to account for the safety barriers HS and their associated functional performance.