Computational Modeling And Design Optimization Under Uncertainty

Associate Professor P. B. Nair 
University of Toronto
Institute for Aerospace Studies
4925 Dufferin St., Ontario, Canada M3H 5T6

Phone: +1-416-667-7720
Fax: +1-416-667-7799
Email: pbn (at) utias.utoronto.ca
Web: arrow.utias.utoronto.ca/~pbn/

Education

  • Ph.D. – University of Southampton
  • M.Tech. – IIT Bombay
  • B.Tech. – IIT Bombay

Awards and Honors

  • Tier II Canada Research Chair in Computational Modeling and Design Optimization Under Uncertainty

Research Overview

Professor Prasanth Nair is the Tier II Canada Research Chair in Computational Modeling and Design Optimization Under Uncertainty and an Associate Professor at UTIAS. He received his Ph.D. (2000) from the University of Southampton, and his M.Tech. (1997) and B.Tech. (1995) degrees in Aerospace Engineering from the Indian Institute of Technology, Mumbai. Prior to joining UTIAS in 2011, he was an academic in the School of Engineering Sciences at the University of Southampton.

Prof. Nair’s research interests lie in three main areas: (i) computational modeling of deterministic and stochastic systems governed by partial differential equations, (ii) optimization algorithms for design, control and parameter estimation, and (iii) generalized function approximation problems. He is the co-author of a book on Aerospace Design (Computational Approaches for Aerospace Design, John-Wiley and Sons, 2005) and over 100 articles in referred journals, edited books and conference proceedings.

Prof. Nair heads the Computational Modeling and Design Optimization Under Uncertainty Group at UTIAS. The research activities of this group are driven by the vision that future computational modelling techniques must not only predict nominal response but also produce a certificate of response variability that rigorously accounts for all sources of uncertainty. Furthermore, this enhanced analysis capability must be highly efficient, parallelizable, and scalable to high-dimensional models. Theoretical and algorithmic advances in these directions are key to realizing the promise of computational models as reliable surrogates of reality as well as enabling robust design optimization of complex real-world systems.

Prof. Nair’s research group also works on various aspects of scientific computing, including the implementation of numerical algorithms on multiprocessor hardware and parallel function decomposition schemes for alleviating the curse of dimensionality encountered in high-dimensional function approximation and solution of parameterized operator equations. Ongoing research projects include:

  • Numerical methods for stochastic partial differential equations;
  • Numerical methods for constructing real-time emulators of high-dimensional engineering systems with applications to robust design optimization and uncertainty analysis;
  • Bayesian methods and greedy algorithms for modelling spatio-temporal datasets and operator problems;
  • Parsimonious design space parameterization strategies;
  • Statistical shape modelling using noisy and sparse measurement data; and
  • Computational methods for robust design of total hip and knee replacements and emulators for pre-clinical decision support.