- Post Doctoral
MIT Unit Affiliation:
- Civil and Environmental Engineering
- Electrical Engineering & Computer Science
Post Doc Sponsor / Advisor:
Date PhD Completed:
Top 3 Areas of Expertise:
I am a post-doctoral associate in the Department of Civil and Environmental Engineering at MIT in Prof. Oral Buyukozturk's LISS group. I have research interests in novel sensor systems, specifically video cameras for quantitative measurement and laser vibrometers, and I work on applying them to structural health monitoring and non-destructive testing problems. I work closely with Prof. Bill Freeman's and Prof. Fredo Durand's groups in CSAIL. I received my Ph.D in Structures and Materials from the Department of Civil and Environmental Engineering at MIT in 2016, an S.M. in Civil and Environmental Engineering from MIT in 2013, and a B.S. in Physics from Caltech in 2009.
Expected End Date of Post Doctoral Position:
Video Camera-based Vibration Measurement
Functional infrastructure – including transportation, energy, and buildings and other facilities – is key to the economic production of a country and the daily lives of its people. Due to deterioration and potential damage, timely inspections and repairs are necessary to keep infrastructure functioning at full capacity. Visual testing is one of the oldest and most widely used methods for condition assessment; however this technique is limited as it is an inspector's subjective rating rather than an objective measure of structural condition.
Quantitative alternatives to visual testing have emerged over time. These include vibration analysis, in which a structure’s operational resonant frequencies and mode shapes are measured and compared against a healthy baseline to detect changes. Typically, contact sensors such as accelerometers, have been used to measure vibrations. When physical access to a structure is limited or the placement of contact sensors is too time-consuming, new technologies that allow for non-contact measurements can be used. Video cameras, where each pixel is effectively a sensor, can remotely collect a large amount of data from a structure. The challenge is then to interpret these videos into quantitative vibration data.
In this research, newly developed computer vision techniques for analyzing small motions in videos, as applied to the vibration measurement and condition assessment of infrastructure are presented. They allow for qualitative visualizations of normally imperceptible motions as a form of enhanced visual testing, and quantitative measurements of the displacements and vibrations of structures as a basis for condition assessment. Computer vision algorithms for processing video are described and the technique is experimentally validated against traditional sensors. The methodology is demonstrated with a series of laboratory measurements on simple representative structures and field measurements of civil infrastructure, including the WWI Memorial Bridge in Portsmouth, New Hampshire.