Education

• AR0303: Computational Intelligence for Integrated Design 

- Probability & statistics

- Optimization 

- Supervised learning 

- Unsupervised learning 

- Reinforcement learning

Link to course page (OPEN ACCESS material)

• AR3B012: Computational Repertoire for Design

- Intro to finite elements 

- Basics of risk assessment

- Design optimization

- GenAI for design

Link to course page (Course guide)

• IFEEMCS520100: Fundamentals of AI Program

- Probabilistic modeling of degrading systems

- Intervention planning formulation

- Optimization methods

- Deep reinforcement learning

Link to course page (Course guide)

Prospective students

Highly motivated students interested in doing their MSc thesis on artificial intelligence, probabilistic, and/or optimization topics for the built environment are welcome to contact me. Examples of past projects can be found below. 

Past students

• Stefan Pol, Modular ribbed floor system: Optimization of the combinatorial problem, Structural Engineering MSc Track, TU Delft.

• Antonios Mavrotas, Q-Res MARL: A resilience-based reinforcement learning tool for post-earthquake repair scheduling of interdependent infrastructures, Building Technology MSc Track, TU Delft.

• Andreas Mananas, Topology optimization for process-induced anisotropy in glass structures, Building Technology MSc Track, TU Delft.

• Emily Lenarduzzi, Building energy meta models: Predicting building energy performance at city scale, Building Technology MSc Track, TU Delft.

• Xiaochen Ding, Closed-loop control of robotic 3D clay printing using machine learning for overhang structure, Building Technology MSc Track, TU Delft.

• Laila Hany Mostafa Awad Saleh,  A methodology for human-AI collaboration in early architectural design: A prototype for generating data driven 3D volumetric models using natural language input, Building Technology MSc Track, TU Delft.

• Elena Macedo Dauzacker, Less stress for glass giants: A stress-based topology optimization workflow for cast glass, Building Technology MSc Track, TU Delft.

• Swornava Guha, 3DP glass assemblies: A novel workflow for circular assembly using 3D-printed glass masonry units with a kirigami inspired interlayer, Building Technology MSc Track, TU Delft.

• Shreya M. Kejriwal, Inspection & maintenance planning for transport infrastructure management using deep reinforcement learning, Building Technology MSc Track, TU Delft.

• Jakub Wyszomirski, ReinforceRay: Optimal long-term planning of photovoltaic and battery storage systems in grid-connected residential sector with reinforcement learning, Building Technology MSc Track, TU Delft. [tutorial]

• Jair Lemmens, Deep generative design: Deep reinforcement learning for performance-based design assistance, Building Technology MSc Track, TU Delft.

• Tahir Ishrat, Generative design of Catalan vaults for multi-storey seismic construction, Building Technology MSc Track, TU Delft.

• Pim Brueren, To3DPGS: Development of an algorithm for topology optimized 3D printed glass structure, Building Technology MSc Track, Architecture & the Built Environment, TU Delft.

• Sasipa Vichitkraivin, Inspection and maintenance planning of historical timber structure concerning climate change implementing reinforcement learning, Building Technology MSc Track, TU Delft.

• Nefeli Karadedou Isoua, Retrofitting planning optimization, Building Technology MSc Track, TU Delft.

• Eva Schoenmaker, Development of a three-dimensional topology optimization algorithm for mass-optimized cast glass components, Building Technology MSc Track, TU Delft.

• Amy Sterrenberg, Deep generative design: A deep learning framework for optimized spatial truss structures with stock constraints, Building Technology MSc Track, TU Delft. [tutorial]

• Lisa-Marie Mueller, 3D generative adversarial networks to autonomously generate building geometry, Building Technology MSc Track, TU Delft. [tutorial]

• Konstantinos Krachtopoulos, Multi-objective Deep Reinforcement Learning for predictive maintenance of road networks, Computer Science MSc Track, TU Delft.

• Sammie Knoppert, Incorporating climate change in I&M planning for engineering structures - A Deep Reinforcement Learning Framework, Structural Engineering MSc Track, TU Delft.

• Yu Tung, Machine learning-assisted analysis of energy consumption profiles and efficiency in Uilenstede campus buildings, Building Technology MSc Track, TU Delft.

• Nathanail Tzoutzidis, Quantification of thermal resilience in buildings: Evaluation of Building Envelope Performance and Operational Parameters, Building Technology MSc Track, TU Delft.

• Stella Pavlidou, Deep Generative Designs: A Deep Learning Framework for Optimized Shell Structures, Building Technology MSc Track, TU Delft. [tutorial] 

• Anna Maria Koniari, Just Glass: Development of a Topology Optimization Algorithm for a Mass-Optimized Cast Glass Component, Building Technology MSc Track, TU Delft. [tutorial]

• Namrata Baruah, Integrated bio-inspired Design by AI: Using cell structure patterns to train an AI model to explore topology design ideas, Building Technology MSc Track, TU Delft. [tutorial]

• Maryamsadat Aboueimehrizi, Machine learning-based assessment tool for predicting daylight and visual comfort, Building Technology MSc Track, TU Delft.

• Christos Lathourakis, Optimal maintenance of deteriorating systems integrating deep reinforcement learning and Bayesian inference, Structural Engineering MSc Track, TU Delft. [tutorial]

• Yogesh Rathod, Evaluation of seismic structural fragility methodologies using non-stationary simulated ground motions, Structural Engineering MSc Track, Penn State University.

• Payal Thukral, Nonlinear programming solvers for hybrid finite element, Structural Engineering MSc Track, Penn State University. 

Course tutorials

• Finite elements & topology optimization with Ansys

• Probability and sampling in engineering problems

• Uncertainty quantification and sensitivity analysis

• Markov chains and Markov decision processes

• Solving MDPs with value iteration

• Solving POMDPs with policy iteration

• Reinforcement learning with Q-learning

• Support vector machines with Scikit-learn and Jupyter 

• Naive Bayes classifier (binary) with Scikit-learn and Jupyter

• Naive Bayes classifier (multi-class) with Scikit-learn and Jupyter

• Multinomial logistic regression with Scikit-learn and Jupyter

• K-nearest neighbor classifier with Scikit-learn and Jupyter

• Neural networks with Scikit learn and Jupyter

• Decision tree classifier with Scikit-learn and Jupyter

• Gaussian mixtures with Scikit-learn and Jupyter

• K-means clustering with Scikit-learn and Jupyter

• Spectral clustering with Scikit-learn and Jupyter

• Self-organizing maps with Scikit-learn and Jupyter

• Surrogate models

• Generative models

Theses tutorials

• Generative structural design through variational autoencoders 1

• Generative structural design through variational autoencoders 2

• Bio-inspired generative design through variational autoencoders

• Vibration-based maintenance planning through reinforcement learning

• Generative design through generative adversarial networks

• Topology optimization for glass structures under manufacturing constraints

• Installation planning of PV systems through reinforcement learning 

Paper tutorials

coming soon.