INAF - Istituto Nazionale di Astrofisica [2024-to present]
ASTRI-SERVIMON: Advanced Monitoring for Astronomical Observatories
This research thread focuses on developing sophisticated monitoring systems for the Cherenkov Telescope Array Observatory (CTAO). The core project, ASTRI-SERVIMON, is an AI-driven platform designed for predictive maintenance and real-time monitoring of complex astronomical infrastructure.
The work aims to enhance the observatory’s alarm and monitoring subsystems in distributed environments. A key innovation involves analyzing historical time-series data from telescopes like the ASTRI-Horn to trace component wear signatures and predict failures before they occur. This represents a shift from reactive to proactive maintenance, creating a comprehensive system for logging, alarms, and monitoring that ensures maximum operational efficiency and minimizes downtime for critical scientific instruments.
Industrial Time Series: Advanced Anomaly Detection for Predictive Maintenance
This thread applies advanced data analysis techniques to industrial contexts, specifically for anomaly detection and predictive maintenance of complex machinery like high-pressure compressors.
The research introduces innovative methodologies that combine time series segmentation with heterogeneous ensemble models. Instead of analyzing data in its entirety, the approach advocates for intelligent segmentation to uncover more subtle, localized anomalous patterns. By then employing an ensemble of different machine learning algorithms, the system leverages the strengths of each model, significantly improving detection accuracy. This philosophy asserts that the complexity of industrial data requires equally sophisticated and nuanced analytical approaches to effectively predict and prevent failures.
Istitut Pasteur [2022-2023]
Viral Metagenomics and Bat Virome Discovery
This research thread focuses on using advanced viral metagenomics to discover and characterize novel viruses in wildlife, with a specific emphasis on bats as key reservoirs for emerging pathogens.
The work is built upon a foundation of methodological innovation, exemplified by the development of two computational pipelines: PIMGAVir and Vir-MinION. These tools are designed for the complete baseline analysis of viral metagenomic data from both 2nd (Illumina) and 3rd (Oxford Nanopore) generation sequencing technologies, enabling comprehensive and efficient virus discovery.
Applying these methods, the research successfully characterized the virome of bat ectoparasites (such as ticks and fleas) in China’s Yunnan Province, a known hotspot for emerging diseases. A significant finding was the identification of novel viruses, including a putative parvovirus and a poxvirus. This discovery is critical for the One Health framework, as it enhances our understanding of the viral diversity circulating in bat populations and their parasites, providing crucial baseline data for early warning and pandemic preparedness by tracing the origins and evolution of potential zoonotic threats.
Harbin Medical University (HMU) [2014-2021]
Computational Biology & Genomics
This research applied bioinformatics and computational methods to diverse biological questions. Key work included analyzing SARS-CoV-2 genome evolution and developing tools for genetic data analysis, such as a package for calculating Hamming distance.
The research also spanned cancer genomics, identifying potential oncogenes in ovarian cancer and studying the anti-metastatic effects of natural compounds. Additional studies investigated bacterial genomics, analyzing genetic patterns in Salmonella to understand its evolution and virulence. This body of work demonstrates the application of data analysis to advance understanding in virology, oncology, and microbiology.
University of Catania [2006-2014]
Computational Methods & Biomedical Applications
This early research focused on developing computational algorithms for biomedical and optimization problems. A major theme was applying mathematical models to cancer research, including using genetic algorithms for therapy optimization and creating personalized schedules for cancer vaccines.
The work also involved developing efficient computational tools, such as HAMFAST for fast Hamming distance calculation and simulated annealing for finding optimal protocols. Additional research explored deploying enterprise services on grid computing systems. This period established a foundation in creating computational solutions for complex biological and technical challenges.