Metasurfaces and Resonators are physical structures that are patterned into crystals to create special quantum-like behavior when it interacts with light; think about how light reflects off a flat CD/DVD disc and creates a rainbow. In this research we study Metasurfaces and Resonators in transparent and reflective crystals to enable new effects which can be used for photonic circuits, waveguides, and sensors. This work doesn’t require prior knowledge in optics but requires a desire to learn new concepts. The research tasks will include:

o   Simulations using Software

o   Data analysis and Statistics

o   Collecting microscopy and optical spectroscopy data

o   Scientific Reading and Writing

Team Terminator’s new project focuses on developing a Chess playing robot that relies on camera inputs to identify a chess board and a mechanical arm to move the chess pieces and on machine learning principles for ideal victory paths for the robot. This project aims to cover various engineering fields and expose participants to long term research in addition to applied and collaborative engineering environments. While a robot that can play a game of chess on its own is the end goal, for this academic year, we will a robot which plays the game of Tic Tac Toe.  In subsequent years, we will gradually incorporate new features to achieve the end goal of building a robot which plays chess. The project will have mechanical, electrical, and programming aspects.

Thermal Comfort Prediction in Smart Buildings via Machine Learning:

Heating, Ventilation, and Air Conditioning (HVAC) is very energy-consuming, accounting for 40% of total building energy consumption in USA. Building thermal control is important for providing high quality working and living environments, because occupants can only feel comfortable when the temperature and humidity of the indoor thermal condition are within the thermal comfort zone.  The key to realize this vision is to predict the accurate thermal comfort model. We will propose a machine learning based approach to learn an individual’s thermal comfort model.  We will use and investigate several algorithms for thermal comfort predictions. Algorithms like artificial neural network (ANN), support vector machine (SVM), Bayesian algorithms, regression, and random forest (RF) classifier. We will use the ASHRAE RP-884 database which is a quality-controlled database collected from field thermal comfort studies in 160 buildings around the world.

The goal of this project is to develop a data analytic and machine learning (reasoning) algorithm to analyze social media data to predict social events such as opioid crisis, hate crime, and social unrest.   Related subjects such as fake information/news detection and radicalization warning are of interests of the project.  Information entropy based classification, clustering, and decision-making is applied to develop the algorithm.  Python-based coding and implementation is planned along with data analysis and visualization with R.

Solar arrays use sunlight as a source of energy to generate DC electricity. Due to the relatively low efficiency of modern day solar cells (~20% maximum) solar powered electronics must be energy efficient for practical use. In the field of embedded systems, some modern day microcontrollers have peripherals and functionalities designed to minimize the amount of energy that is needed to interface with and control systems, making them perfect for solar powered projects. The goal of this project is to design and build a solar powered vehicle that can efficiently maximize solar array energy to charge batteries, as well as minimizing the energy lost through microcontroller control power consumption.