Francesca Cairoli
Portrait of Francesca Cairoli

Tenure-Track Assistant Professor (RTT)

Francesca Cairoli

Department of Mathematics, Informatics and Geoscience
University of Trieste, Italy

My research focuses on the safety assurance of autonomous systems, strategically integrating the scalability of machine learning with the rigorous, verifiable guarantees of formal methods. I combine scientific machine learning and neuro-symbolic AI to embed structured domain knowledge and safety constraints directly into the learning process — using deep generative models to reason about complex, stochastic systems and developing efficient, statistically grounded quantifications of uncertainty for safety-critical applications.

✉ Email Google Scholar Scopus ORCID AI Lab @ UniTS

News

  • Dec. 2026 · 🌺 Honolulu, USA

    🎉 Paper accepted at CDC 2026!

    "Certifiably Robust Reinforcement Learning under Spatio-Temporal Logic Specifications" is heading to the International Conference on Decision and Control (CDC) in Honolulu, Hawaii.

  • Oct. 2026 · 🇮🇹 Trieste, Italy

    🏆 Research funding awarded by Fondazione AgorAI!

    Awarded a two-year postdoctoral research fellowship supported by Fondazione AgorAI for the project "Interpretable Pattern Extraction and Runtime Monitoring for Generative Models". Role: Principal Investigator.

  • Sep. 2026 · 🍕 Naples, Italy

    🎉 Paper accepted at CAESAR Workshop @ ECML-PKDD!

    "Localized Anomaly Detection via Differentiable D-vine Copulas" is off to the CAESAR Workshop at ECML-PKDD 2026 in Naples.

  • Jul. 2026 · 🇵🇹 Lisbon, Portugal

    🎉 Paper accepted at KR 2026!

    "Time Robustness for Point-Based Semantics of Metric Interval Temporal Logic" accepted at the International Conference on Principles of Knowledge Representation and Reasoning (KR) in Lisbon.

  • May 2026 · 🏝️ Paphos, Cyprus

    🎉 Paper accepted at AAMAS 2026!

    "Guiding Neuro-Symbolic Scenario Generation with Spatio-Temporal Logic" accepted at the International Conference on Autonomous Agents and Multiagent Systems (AAMAS) in Paphos.

  • Mar. 2026 · 🇮🇹 Trieste, Italy

    🚀 Promoted to Tenure-Track Assistant Professor!

    Officially an RTT at the Dept. of Mathematics, Informatics and Geoscience, University of Trieste.

  • Sep. 2025 · ⛰️ Graz, Austria

    🎉🎉 Double acceptance at RV 2025!

    "Conformal Predictive Monitoring for Multi-Modal Scenarios" and "CoCAI: Copula-based Conformal Anomaly Identification for Multivariate Time-Series" — both accepted at the same conference.

Research Interests

The core of my approach leverages scientific machine learning and neuro-symbolic AI to embed structured domain knowledge and safety constraints directly into the learning process. This principled integration is designed to significantly enhance model robustness, generalisation, and interpretability. Key directions involve employing deep generative models to analyse and reason about complex, stochastic systems, coupled with the development of efficient, statistically grounded quantifications of uncertainty. These methodological foundations are crucial for certifying reliable and trustworthy performance in safety-critical applications.

Safety AssuranceAutonomous SystemsMachine Learning Formal MethodsGenerative AIStochastic Systems Scientific Machine LearningNeuro-Symbolic AIUncertainty Quantification

Academic Employment

  • From March 2026

    Tenure-Track Assistant Professor (RTT)

    S.S.D. INFO-01/A — Dept. of Mathematics, Informatics and Geoscience, University of Trieste

  • Mar. 2023 – Feb. 2026

    Junior Assistant Professor (RTD-A)

    S.S.D. INFO-01/A — Dept. of Mathematics, Informatics and Geoscience, University of Trieste
    Project: EU PNRR iNEST — Spoke 9, Models, Methods, Computing Technologies for Digital Twin

  • Mar. 2022 – Feb. 2023

    Research Fellowship

    Dept. of Mathematics and Geoscience, University of Trieste
    Project: PRIN 2017 SEDUCE — Machine Learning for the Verification and Synthesis of Cyber-Physical Systems
    Supervisor: Luca Bortolussi

  • Jul. 2017 – Jul. 2018

    Research Fellowship

    Dept. of Engineering and Architecture, University of Trieste
    Project: POR FESR 2014-2020 with Generation Byte — MDCLS: Dynamical Models for Multi-Device Closed-Loop System
    Supervisor: Felice Andrea Pellegrino

Education

  • Nov. 2018 – Oct. 2022

    PhD in Computer Science

    “Earth Science, Fluid-Dynamics and Mathematics. Interactions and Methods”, University of Trieste
    Thesis: Deep Learning for Abstraction, Control and Monitoring of Complex Cyber-Physical Systems
    Supervisor: Luca Bortolussi

  • Oct. 2014 – Mar. 2017

    Master Degree in Mathematics

    University of Trieste
    Thesis: Non-linear whole-genome analysis of DNA methylation fidelity
    Supervisor: Luca Bortolussi

  • Sept. 2010 – Jul. 2013

    Bachelor Degree in Mathematics

    University of Milano-Bicocca
    Thesis: Applicazioni computazionali delle basi di Groebner su equazioni polinomiali
    Supervisor: Andrea Previtali

Research & Study Visits

  • University of Southern California, CPS-Vida Lab, Los Angeles, USA — May–Aug. 2023
  • University of Konstanz, Centre for the Advanced Study of Collective Behaviour, Germany — Feb.–Mar. 2022
  • Royal Holloway University, London, UK — Jan.–Feb. 2020
  • Saarland University, Dept. of Modeling and Simulation, Germany — Oct. 2016
  • Graz University of Technology, Austria — Erasmus, Sept. 2012–Mar. 2013

Publications

  1. Cairoli, F.; Bortolussi, L. Certifiably Robust Reinforcement Learning under Spatio-Temporal Logic Specifications. International Conference on Decision and Control (CDC), 2026.
  2. Pearson, N. A.; Zanello, F.; Russo, D.; Bortolussi, L.; Cairoli, F. Localized Anomaly Detection via Differentiable D-vine Copulas. CAESAR Workshop @ ECML-PKDD, 2026.
  3. Silvetti, S.; Compagnucci, I.; Cairoli, F.; Trubiani, C.; Nenzi, L. Time Robustness for Point-Based Semantics of Metric Interval Temporal Logic. International Conference on Principles of Knowledge Representation and Reasoning (KR), 2026.
  4. Bonin, L.; Giacomarra, F.; Bortolussi, L.; Deshmukh, J.; Cairoli, F. Guiding Neuro-Symbolic Scenario Generation with Spatio-Temporal Logic. International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2026. DOI
  5. Cairoli, F.; Bortolussi, L. Scalable and reliable stochastic parametric verification with stochastic variational smoothed model checking. International Journal of Systems Science, 1–29, 2025. DOI
  6. Cairoli, F.; Kuipers, T.; Bortolussi, L.; Paoletti, N. Conformal quantitative predictive monitoring of stochastic systems with conditional validity. Nonlinear Analysis: Hybrid Systems, 57, 101606, 2025. DOI
  7. Pearson, N. A.; Zanello, F.; Russo, D.; Bortolussi, L.; Cairoli, F. CoCAI: Copula-based Conformal Anomaly Identification for Multivariate Time-Series. International Conference on Runtime Verification (RV), 2025. DOI
  8. Cairoli, F.; Bortolussi, L.; Deshmukh, J.; Lindemann, L.; Paoletti, N. Conformal Predictive Monitoring for Multi-Modal Scenarios. International Conference on Runtime Verification (RV), 2025. DOI
  9. Bortolussi, L.; Cairoli, F.; Klein, J.; Petrov, T. Neuro-Symbolic Discovery of Markov Population Processes. International Conference on Neuro-symbolic Systems (NeuS), PMLR, 2025. Link
  10. Giacomarra, F.; Hosseini, M.; Paoletti, N.; Cairoli, F. Certified Guidance for Planning with Deep Generative Models. International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2025. Link
  11. Randone, F.; Doz, R.; Cairoli, F.; Bortolussi, L. Towards a probabilistic programming approach to analyse collective adaptive systems. International Symposium on Leveraging Applications of Formal Methods (ISoLA), 2024. DOI
  12. Cairoli, F.; Anselmi, F.; d'Onofrio, A.; Bortolussi, L. Generative abstraction of Markov population processes. Theoretical Computer Science, 977, 114169, 2023. DOI
  13. Anselmi, F.; Manzoni, L.; D'onofrio, A.; Rodriguez, A.; Caravagna, G.; Bortolussi, L.; Cairoli, F. Data symmetries and learning in fully connected neural networks. IEEE Access, 11, 47282–47290, 2023. DOI
  14. Cairoli, F.; Bortolussi, L.; Paoletti, N. Learning-Based Approaches to Predictive Monitoring with Conformal Statistical Guarantees. International Conference on Runtime Verification (RV), 2023. DOI
  15. Bortolussi, L.; Cairoli, F.; Carbone, G.; Pulcini, P. Scalable Stochastic Parametric Verification with Stochastic Variational Smoothed Model Checking. International Conference on Runtime Verification (RV), 2023. DOI
  16. Bortolussi, L.; Cairoli, F.; Giacomarra, F.; Scassola, D. Model Abstraction and Conditional Sampling with Score-Based Diffusion Models. International Conference on Quantitative Evaluation of Systems (QEST), 2023. DOI
  17. Bortolussi, L.; Cairoli, F.; Klein, J.; Petrov, T. Data-Driven Inference of Chemical Reaction Networks via Graph-Based Variational Autoencoders. International Conference on Quantitative Evaluation of Systems (QEST), 2023. DOI
  18. Cairoli, F.; Paoletti, N.; Bortolussi, L. Conformal Quantitative Predictive Monitoring of STL Requirements for Stochastic Processes. 26th ACM International Conference on Hybrid Systems: Computation and Control (HSCC), 2023. DOI
  19. Ballarin, E.; Bortolussi, L.; Cairoli, F.; Gallese, C.; Nenzi, L.; Saveri, G. Reliable and Explainable AI in Trieste. Ital-IA, 394–396, 2023. PDF
  20. Cairoli, F.; Paoletti, N.; Bortolussi, L. Neural predictive monitoring for collective adaptive systems. International Symposium on Leveraging Applications of Formal Methods (ISoLA), 30–46, 2022. DOI
  21. Bortolussi, L.; Cairoli, F.; Paoletti, N.; Smolka, S. A.; Stoller, S. D. Neural predictive monitoring and a comparison of frequentist and Bayesian approaches. International Journal on Software Tools for Technology Transfer, 1–26, 2021. DOI
  22. Cairoli, F.; Bortolussi, L.; Paoletti, N. Neural predictive monitoring under partial observability. International Conference on Runtime Verification (RV), 121–141, 2021. DOI
  23. Cairoli, F.; Carbone, G.; Bortolussi, L. Abstraction of Markov Population Dynamics via Generative Adversarial Nets. International Conference on Computational Methods in Systems Biology (CMSB), 19–35, 2021. DOI
  24. Bortolussi, L.; Cairoli, F.; Carbone, G.; Franchina, F.; Regolin, E. Adversarial Learning of Robust and Safe Controllers for Cyber-Physical Systems. IFAC-PapersOnLine 54(5), 223–228, 2021. DOI
  25. Cairoli, F.; Fenu, G.; Pellegrino, F. A.; Salvato, E. Model Predictive Control of Glucose Concentration Based on Signal Temporal Logic Specifications with Unknown-Meals Occurrence. Cybernetics and Systems, 51(4), 426–441, 2020. DOI
  26. Bortolussi, L.; Cairoli, F.; Paoletti, N.; Smolka, S. A.; Stoller, S. D. Bayesian Neural Predictive Monitoring. 2nd Workshop on AI and Formal Verification, Logics, Automata and Synthesis (OVERLAY), 95–100, 2020. PDF
  27. Bortolussi, L.; Cairoli, F. Bayesian abstraction of Markov population models. International Conference on Quantitative Evaluation of Systems (QEST), 259–276, 2019. DOI
  28. Cairoli, F.; Fenu, G.; Pellegrino, F. A. Clinical Decision Support Using Colored Petri Nets: a Case Study on Cancer Infusion Therapy. 6th International Conference on Control, Decision and Information Technologies (CoDIT), 314–319, 2019. DOI
  29. Cairoli, F.; Fenu, G.; Pellegrino, F. A.; Salvato, E. Model predictive control of glucose concentration based on signal temporal logic specifications. 6th International Conference on Control, Decision and Information Technologies (CoDIT), 714–719, 2019. DOI
  30. Bortolussi, L.; Cairoli, F.; Paoletti, N.; Smolka, S. A.; Stoller, S. D. Neural predictive monitoring. International Conference on Runtime Verification (RV), 129–147, 2019. DOI
  31. Bortolussi, L.; Cairoli, F.; Paoletti, N.; Stoller, S. D. Conformal predictions for hybrid system state classification. From Reactive Systems to Cyber-Physical Systems, 225–241, 2019. DOI

Full and up-to-date list on Google Scholar.

Teaching

  • Computer Science — Degree in Medicine and Surgery, 2026/27
  • Data-Driven System Engineering — MSc in Computer Engineering, 2026/27
  • Scientific Machine Learning and Simulation Intelligence — MSc in Data Science and AI, 2026/27
  • Explainable, Causal and Neuro-Symbolic Artificial Intelligence — MSc in Data Science and AI, 2026/27
  • Explainable and Neuro-Symbolic Artificial Intelligence — MSc in Data Science and AI, 2025/26
  • Symbolic and Neuro-Symbolic Artificial Intelligence — MSc in Data Science and AI / Scientific and Data Intensive Computing, 2024/25
  • Simulation Intelligence and Learning for Autonomous Systems — MSc in Data Science and AI / Scientific and Data Intensive Computing, 2023/24–2025/26
  • Computer Programming (preparatory course) — MSc in Data Science and AI / Scientific and Data Intensive Computing, 2023/24–2025/26
  • Stochastic Modeling and Simulation (Teaching Assistant) — MSc in Data Science and Scientific Computing, 2018/19–2021/22

Grants

  • Fondazione AgorAI postdoctoral research fellowship (2-year) for “Interpretable Pattern Extraction and Runtime Monitoring for Generative Models”. Role: Principal Investigator.
  • iNEST Young Researchers grant (Mar. 2024 – Oct. 2025) for “Reactive Scenario Generation for Testing Resilience of Autonomous Systems (RESGEN)”. Role: Principal Investigator.

Research Projects

  • iNEST — Interconnected Nord-Est Innovation Ecosystem, EU NextGenerationEU / PNRR, Spoke 9. Collaboration: Univ. Padova, Univ. Trieste, OGS, SISSA. Role: Research Topic Co-Coordinator.
  • SEDUCE — Designing Spatially Distributed Cyber-Physical Systems under Uncertainty, PRIN 2017. Collaboration: IMT Lucca, GSSI L'Aquila, Univ. Trieste, Univ. Camerino/Firenze.
  • MDCLS — Multi Device Closed Loop System, POR FESR 2014-2020. Collaboration: Generation Byte Srl, Glance Vision Tech Srl, Univ. Trieste.

Academic Supervision

Postdocs

  • Former Francesca Randone: Probabilistic programs as stochastic surrogates (RESGEN project)

PhD Students

  • Current Nicholas Pearson (ADSAI PhD, UniTS): Time series analysis applied to the urban water cycle
  • Current Alessandro Della Siega (ADSAI PhD, UniTS, Co-supervision): Discovery of Stochastic Differential Equations
  • Former Francesco Giacomarra (ADSAI PhD, UniTS, Co-supervision): Certification of Deep Generative Models
  • Former Julia Klein (Konstanz University, Co-supervision): Data-Driven Inference of Chemical Reaction Networks

BSc and MSc Students

  • Former Alessandro Della Siega: Identification of stochastic processes (MSc, DSAI)
  • Former Annalisa Paladino: Gradient-based optimisation of spatio-temporal logical objectives (Internship & MSc, DSAI)
  • Former Mattia Simonutti: Conformal Inference as a Regularisation Term in Regression Models (BSc, 2024)
  • Former Matteo Munini: Theoretical modeling of emergency events: Social Forces in Extreme Conditions (MSc, 2025)
  • Former Davide Bartole: Model abstraction using Temporal Difference VAEs (MSc, 2022)
  • Former Gaia Saveri: Graph Neural Networks for Propositional Model Counting (MSc, 2021)
  • Former Matteo Boi: Adversarial Learning on Model of Artificial Pancreas (BSc, 2021)
  • Former Francesco Franchina: Adversarial Learning of Robust and Safe Controllers for CPS (MSc, 2020)
  • Former Laura Falciani: Bayesian Abstraction for Chemical Reaction Networks (MSc, 2019)

Beyond the Lab

Outside the lab, I'm happiest in the mountains — climbing, hiking, and chasing new adventures.

Traversing a rock face above a green valley
Climbing rope against a rock wall
Reaching for the next hold while climbing
Standing on a mountain ridge summit

Contact

Department of Mathematics, Informatics and Geoscience — University of Trieste, Italy
Via Giovanni e Demetrio Economo, 12/3 — Room 415

✉ francesca.cairoli@units.it AI Lab @ UniTS Google Scholar Scopus ORCID