We are a group focusing on application of statistical and physical theories, models and simulation techniques to better quantify the quality, reliability, variability and uncertainty in performance / lifetime of any product / system / service with wide applications ranging from the nano-world of atoms and thin films all the way to the macro world of power grids, manufacturing and transportation systems and even system-of-systems (SoS). We model, optimize, manage and re-design the product / system to ensure robustness in the “design” phase.

Our ultimate goal is to provide an open source “design for reliability (DFR) toolkit” that can be used by the design community to aid in their activities across several industries. The toolkit will encompass several of the following features, but not limited to these as such.

  • Structural abstraction of system in the form of a network / block diagram.
  • Behavioral abstraction of system by modeling interactions of different units using physical knowledge or by empirical means.
  • Monte Carlo based code for simulating uncertainty and evaluation of worst-case scenarios.
  • Use of analytical models (Bayesian inference) and machine learning techniques (Neural Networks) as well as optimization routines (Genetic algorithms, Simulated annealing) to better quantify the estimated failure probabilities and manage resources with risk incorporated.
  • Development of advanced predictive analytics techniques that enable early inference of remaining useful life (RUL), banking on real-time sensor data feeds, for prior planning of inventory and maintenance schedules to minimize unanticipated downtimes and production losses.

Our studies have implications on the entire paradigm of a product’s lifecycle ranging from its design, manufacturing (quality), usage (reliability), maintenance (warranty) and eventual replacement. Current technologies of focus (as test beds for our algorithms) include :-

  • Additive manufacturing (3D printing).
  • Self-assembled growth of nanostructures.
  • Thin Films in Nanoelectronic Devices.
  • Prognostics and Health Management of Industrial Systems.
  • Robustness study of system-of-systems (SoS).

The nature of research in our group is highly multi-disciplinary transcending the fields of physics, materials science, statistics, signal processing, optimization and machine learning. If you are interested in career opportunities, research positions or internships, please refer to the “Openings” page for more details.