A low computationally complex method for event classification in Non-Intrusive Load Monitoring systems capable of running on a single chip computer (Pi 3). Presented at 2017 IEEE Massachusetts Institute of Technology Undergraduate Research Technology Conference.
A low cost standalone system for performing device identification and monitoring power consumption in a rural household in India. Presented at 2016 IEEE Dallas Circuits and Systems Conference.
A Machine Learning model that predicts the occupancy of a movie using its booking history. (Done in collaboration with one of the top three multiplex chains in India).
A slack-bot that displays students’ attendance and progress when summoned by commands.
Used : Slacker, Python, Pandas, Google API, Heroku
Assembled and configured a Server with NAS dedicated for Research Purposes at SF.
Built a high-speed Data aquisition system for building custom NILM dataset.
Used: Beaglebone Black Programmable Real-time Unit (PRU)
A P-controller based algorithm that maintains the speed of the motor at the user given speed.