Tools, products and research we’re proud of.

Free Grasshopper plugin: ML, non-linear optimization, tensor-field manipulation, isosurfaces and linear algebra.

• Machine learning
• Optimization
• Tensor fields
• Isosurfaces
• Linear algebra

Grasshopper plugin for advanced meshing built on CGAL and libigl; quad remeshing from conjugated vector fields.

• CGAL
• libigl
• Quad remeshing
• Vector fields
• C++

Open-source Lean 4 library: tensors, kinematics, stress, constitutive laws, elastostatics, beams and plates, plus a formalised 2013 paper. No sorry, no axioms.

• Lean 4
• Mathlib
• Formal verification
• Continuum mechanics
• Beams & plates
• Homogenization

Greco, Lebée and Douthe (IASS 2013): the square nexorade homogenized as a plate, with closed-form stiffness and an optimal cell geometry. Every result is now a Lean theorem.

• Reciprocal frames
• Nexorades
• Homogenization
• Plates
• Lean 4

Greco, Bevilacqua and Croce (AAG 2014): a funnel-shaped glass and steel roof for the Scuola Normale Superiore courtyard, from form finding to panel moulds and sized beams.

• Computational architecture
• Free-form roofs
• Glass panelization
• Structural optimisation
• Grasshopper
• ABAQUS

Greco (IASS 2018): quad-only gridshells whose edges follow principal stresses and principal curvatures at once, built on CGAL and libigl in the Capybara plugin.

• Gridshells
• Quad remeshing
• Stress fields
• Curvature
• CGAL
• libigl
• Grasshopper

Greco (IASS 2018): machine-learning models trained on Eurocode 3 designs that predict steel end-plate connections, tested on a real London project.

• Machine learning
• Optimization
• Steel connections
• Eurocode 3
• scikit-learn

Margnelli, Greco, Gkoktsi et al. (2018): nine months of full-scale monitoring on a slender 47-storey tower in London, with damping and response identified from the data.

• Wind engineering
• Tall buildings
• Structural monitoring
• System identification
• Signal processing

Gonzalez-Fernandez, Macdonald, Titurus, Greco, Zanchetta et al. (2019): identifying aeroelastic effects on a tall building from full-scale data.

• Wind engineering
• Aeroelasticity
• System identification
• Tall buildings

Field instrumentation of a high-rise: Arduino sensing of wind excitation and tower acceleration, statistical analysis of ~660M data points, and bespoke apps deriving transfer functions and correlations in real time.

A two-stage algorithm that first reduces the curvature of doubly-curved glass panels, then clusters them to minimise the number of moulds — cutting cost and wastage on free-form façades.

Machine-learning and optimization algorithms that suggest steel-joint connections with high accuracy, wrapped in an interactive design framework.

Homogenization studies of the elastic behaviour of reciprocal systems and computational architecture work, presented at IASS 2013 and Advances in Architectural Geometry 2014. The 2013 paper is now formally verified in Lean.