Physics-based models are commonly used to numerically simulate multiphysics dynamical interaction. However, developing accurate physics-based models is not always feasible or straight forward. In such situations, are data-driven models a viable alternative? How robust are data-driven models in representing complex multiphysics interaction, as in the case of a jet engine vibration? My research focuses on time domain and frequency domain data-driven algorithms like dynamic mode decomposition, vector fitting, etc.

How do engineer our structures to automatically detect damages and changes in their behavior? How to identify operational behavioural changes from the damages? To answer them above question one needs to understand the importance of sensing. Sensing is an important parameter across all fields of monitoring, be it human health monitoring, be it engineering structures like bridges, airplanes etc. Sensing is only one part of the problem , the real challenge lies in developing efficient algorithms which can in fact work on the data from the sensors ,preferably real time and detect changes. Working for my master's thesis, we are striving to engineer certain real-time algorithms which can process the information from sensors and detect damages real time. 

Whatever be the field one is working on right now, coding becomes an inevitable part of modern day research . Efficient algorithms call for efficient codes which use minimal time to process the data. Matlab has been extensively used throughout my works which include assignments, TA works, thesis works etc. Other programming languages which I am well versed with are Java, Python etc. NetBeans IDE is the environment which I prefer for Java programs. Have worked on android app development during my undergraduate days.