Hi, I’m Luca

PhD Candidate & Research Assistant
Imperial College London

I use geometric deep learning for learned simulation of physical and engineering systems, from composite failure to cognitive digital twins, knowledge graphs and VR.

Luca Patrignani

Publications

2026

Graph neural networks with hybrid local-global attention for effective prediction of mechanical response in structures

L. Patrignani, S.T. Pinho

Computer Methods in Applied Mechanics and Engineering

2026 · Under review

Autoregressive graph transformers with local-global attention for progressive damage evolution in composite laminates

L. Patrignani, S.T. Pinho

Preprint on SSRN

2026

A structured virtual reality laboratory methodology for teaching structural mechanics to undergraduate engineers

L. Patrignani, D.T. Venetsanos, S.T. Pinho

Computer Applications in Engineering Education

2025

SpaceKG: Towards exploiting knowledge graphs in space systems

L. Patrignani, E. Laurenza, E. Sallinger, A. Vlad, P. Gaudenzi

Acta Astronautica

Talks & Posters

  1. 2025 NASA Langley Research Center Invited seminar · Hampton, USA
  2. 2026 AIComp 2026 ★ Best Paper Award Oral · Vancouver, Canada
  3. 2026 ECCM22 Oral · Oslo, Norway
  4. 2026 Aeronautics Research Showcase ★ First Prize Poster · London, UK
  5. 2025 EUROMECH Colloquium 656 Oral · Gothenburg, Sweden
  6. 2025 COMPOSITES 2025 (ECCOMAS) Oral · Vienna, Austria
  7. 2025 AIComp 2025 Oral · Bristol, UK
  8. 2025 ICCM24 Oral · Baltimore, USA

CV

Experience

  • 2024–nowPhD Researcher & Research AssistantImperial College London
  • 2023–2024Structural AnalystAirbus, Madrid
  • 2022–2023Research CollaboratorBank of Italy, Rome
  • 2021Research CollaboratorNational Institute for Nuclear Physics, Frascati

Education

  • 2024–nowPhD, Advanced Numerical Methods for Aeronautical StructuresImperial College London
  • 2023–2024Master, Business Management & Data AnalyticsUniversidad Camilo José Cela
  • 2021–2023MSc, Space & Astronautical EngineeringSapienza University of Rome
  • 2018–2021BSc, Aerospace EngineeringSapienza University of Rome

Honours

  • 2026Isambard-AI compute grantUK AI Research Resource
  • 2026Best Paper AwardAIComp 2026
  • 2026First Prize, poster competitionAeronautics Research Showcase
  • 2025Best Graduate Teaching AssistantImperial College London

Get in touch

“With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.”

John von Neumann. Graph transformers need slightly more.