Research


BITS Laboratory (Breakthrough Interactive Thinking Systems Lab) is engaged in the research of cyber physical systems that enable novel applications such as ones that involve brain-computer interfaces, medical decision support, and smart cities, among many others. In almost all of these systems the common theme is the application of various Machine Learning (ML) techniques to the acquired – and pre-processed – data to produce a feedback for what the machine "thinks." Based on the application, this feedback can involve i) the training of a human to improve certain skills (brain-computer interfaces), ii) a novel visualization of the data – or suggested probability of certain diseases – presented to medical professionals (medical decision support), or iii) the control of embedded actuators intelligently (smart cities). While the generalized feedback (or feedforward) loop of these systems remains fairly consistent across these applications, the utilized sensors, as well as the pre-processing methods, could change significantly based on the characteristics of the sensors; while brain-machine applications use EEG sensors that are highly susceptible to noise, thereby introducing SNR-related challenges, sensors deployed throughout a city may face different challenges related to the transmission of the acquired data. On the "processing" front, the selection of the proper ML algorithm introduces non-trivial challenges to balance algorithmic performance vs. accuracy. Furthermore, the combination of these selections introduces a co-design challenge. BITS Laboratory, along with its highly-qualified team of PhD, MS, and undergraduate students, is dedicated to researching and developing systems that provide breakthrough solutions to these highly-sophisticated applications.