Predictive Science Lab

Creating artificial intelligence (AI) technologies that accelerate the pace of engineering innovation.

We are an interdisciplinary research group at the School of Mechanical Engineering, Purdue University led by Prof. Ilias Bilionis.

Our mission is to create artificial intelligence (AI) technologies that accelerate the pace of engineering innovation.

Our applications span the range between purely technical (e.g., electric machines, high-performance materials) and sociotechnical (e.g., smart buildings, extra-terrestrial habitats).

Our research develops communication channels between theories and data. The communication protocol is based on probability theory (thought of as an extension of logic under uncertainty) with an additional layer of causality (expressed through physical laws and graphical models) and is realized using modern machine learning techniques based on deep neural networks.

We are extremely grateful for the support of our academic, government, and private sponsors:


News, thoughts, etc.

Making publication-ready figures for papers

This is a quick guide for making figures that are publication-ready using matplotlib. Note: PSL students must folow these directions to …

Meet the Team

Click here to see how past group members are doing.

Principal Investigator


Ilias Bilionis

Associate Professor of Mechanical Engineering

Grad Students


Alana Lund

Ph.D. Candidate


Sharmila Karumuri

Ph.D. Student


Alex Alberts

Ph.D. Student


Vanessa Kwarteng

Ph.D. Student


Andres Beltran

Ph.D. Student


Atharva Hans

Ph.D. Student


Nimish Awalgaonkar

Ph.D. Student



Computing Contact Problems with Self-Conforming Hybrid Materials

Realize a novel hybrid material that self-conforms around an object of interest as a physical route for computing and reporting the …

Resilient Extraterrestrial Habitats

Developing smart autonomous deep space habitats that will adapt, absorb and rapidly recover from disruptions.

Using data science to discover materials with extreme properties.

Discovery of high-temperature, oxidation resistant, complex, concentrated alloys via data science driven multi-resolution experiments …

Physics-Informed LeaRnIng for Multiscale Systems (PILgRIMS)

Baking known physics into machine learning algorithms.

Sociotechnical Systems to Enable Smart and Connected Energy-Aware Residential Communities

Incentivizing subsidized communities to consume less energy.

Automated Decision Support to Address Community Resilience Challenges

Improving the resilience of the built environment through an automated vision-based system that generates probabilistic predictions.

Design of electric machines with manufacturing uncertainties

Development of computational tools for electric machine design including manufacturing uncertainties.

Causal inference on Bayesian graphical networks

Learn potential causal structures with quantified uncertainties from observational data.

A Theoretical Framework for Understanding Strategic Behavior in Systems Engineering

Understanding how incentive misalignment affects the design of complex systems.

Human-centered Systems for Cyber-enabled Sustainable Buildings

Understanding thermal and visual comfort in high-performance buildings.


This is a list of the courses Prof. Bilionis teaches on a regular basis. He has also delivered special short courses on research topics, e.g., uncertainty quantification, Gaussian process regression, global optimization of expensive functions, sequential design of simulations/experiments, which will be added to the site gradually.


Purdue ME 270 – Basic Mechanics I (Statics)

Introduction forces, moments, and static equilibrium.

Purdue ME 274 – Basic Mechanics II (Dynamics)

Introduction to dynamics of rigid bodies

Introduction to Uncertainty Quantification

Quantifying uncertainties in physical systems.