Physics-informed Neural Networks for Time-Dependent Transport Equations

Companies can realize the cost-saving benefits of having access to large amounts of data provided they have digital tools that allow efficient and accurate extraction of information from the dataset. The goal of this project is
to conduct the research required for machine learning algorithms to provide reliable engineering predictions for industrial applications. To achieve the stated objectives, a diverse team of researchers will investigate several
fundamental questions about the interplay between neural networks and other simulation technologies currently used by the industrial sector. With its mandate to develop state-of-the-art tools for physics-based engineering
applications, SOTAES is well-positioned to develop an innovative product based on the outcomes of this research.

Faculty Supervisor:

Mohammad Hassanzadeh

Student:

Partner:

SOTAES Inc.

Discipline:

Engineering

Sector:

Professional, scientific and technical services

University:

University of Windsor

Program: