The GREYDIENT consortium gathered in person for the fourth time since the start of the project. This time, the event was held in Milan at the headquarters of the POLIMI’s Energy Department, and the LASAR team was in charge of the organization.
During the first two days, the ESRs shared updates about their work and got insightful feedback from their colleagues. Leonardo Miqueles (ESR 5) presented about grey-box digital twins and their applications to the risk monitoring and control of nuclear power plants. Juan-Pablo Futalef (ESR 6) explained how to “greyfy” cyber-physical systems and generate scenarios efficiently.
Besides the updates, the ESRs attended trainings on topics such as multi-agent simulations, sensitivity analysis, Bayesian networks, and design thinking. Professionals in charge of these trainings were Prof. Gyunyoung Heo (Kyung Hee Uni.) , Prof. Stefano Magistretti (POLIMI), Prof. Emanuele Borgonovo (Bocconi Uni.), and Prof. Ibrahim Ahmed (POLIMI).
The ESRs also shared experiences and brought up ideas about improving collaboration and the overall experience within the network… and visited the National Museum of Science and Technology Leonardo Da Vinci!
The next GREYDIENT event will be in Thessaloniki, Greece, and Aristotle University of Thessaloniki is in charge of the organization.
With China the largest community in LASAR, the Chinese New Year was everything but quiet. Everyone gathered to celebrate the upcoming year of the Rabbit by bringing a collection of traditional Chinese plates.
This small break was a great chance to exchange smiles and reflect about next events in our group, which is constantly evolving.
Indeed, the year of the Rabbit is expected to bring quietness, yet powerful reflections after times of contemplation.
He won the PhD award sponsored by PayPal Ireland at the ESREL2022 conference, Dublin, 28th August – 1st September 2022.
The work “A Taxonomy for Modelling Reports of Process Safety Events in the Oil and Gas Industry” has received the award in the category “Innovation in Energy Sector Applications”. It proposes an innovative methodology for quantitative risk assessment, which combines Natural Language Processing with Bayesian Networks for the systematic analysis of textual reports in hydrocarbon production assets.
The developed work is part of the multidisciplinary PhD program between the Energy Department and Department of Electronics, Information and Bioengineering of Politecnico di Milano and it has been developed within the ongoing PhD research of Dario Valcamonico under the supervision of Prof. Piero Baraldi and Prof. Enrico Zio in collaboration with our industrial partners Anna Crivellari, Luca Decarli and Laura La Rosa from Eni Natural Resources.
In multi-component systems, degradation, maintenance, renewal and operational mode change continuously the operating conditions. The identification of the onset of abnormal conditions from signal measurements taken in such evolving environments can be quite challenging, due to the difficulty of distinguishing the real cause of the signal variations. In this work, we present a method for fault detection in evolving environments that uses a Sparse Autoencoder-based Deep Neural Network (SAE-DNN) and a novel procedure that remarkably reduces the computational burden for setting the values of the hyperparameters. The method is applied to a synthetic case study and to a bearing vibration dataset. The results show that it is able to accurately detect faults in multi-component systems, outperforming other state-of-the-art methods.