Xerox Engineering Research Fellows
2020 Research Opportunities
Electrical and Computer Engineering
Research Project: Design and Optimization of Large Scale Software Defined Wireless Networks
Performing communication and computation in an ad hoc network of mobile devices is challenging yet critical for next-generation wireless networks. To ensure that data can be efficiently communicated where it is needed, when it is needed, we are working on the development of technologies and approaches for achieving robust data connections in heterogeneous network platforms using a mixture of ad hoc and hierarchical networks.
In order to support robust and efficient communication in a hybrid infrastructure/mobile ad hoc network, we are working on techniques to 1) automatically initialize and maintain connectivity in a multi hop ad hoc network; 2) extend connectivity by deploying UAVs that provide off-loading of data from a multi-hop ad hoc network to the infrastructure; and 3) select and optimize network protocol parameters to improve the overall network performance.
The work conducted in this project combines theoretical mathematical modeling and optimization with practical implementation of the developed protocols on emulation/simulation systems as well as a Raspberry Pi testbed to verify performance.
Research Project Descriptions
Research Project #1: Developing New Topological Electronics from Quantum Materials
A remarkable number of new topologically non-trivial quantum electronic materials have been identified recently. These materials offer new types of functionality and controllability that go well beyond what conventional electronics can do. Exploiting these new degrees of freedom can lead to significant advancements for the electronics industry as Moore's law comes to a close. In our lab we will be focusing on exploring the dissipationless chiral transport properties of these materials to generate topological transistors. Students will work in multiple settings, learning both microfabrication techniques in the cleanroom, as well as materials growth and characterization in our materials synthesis lab. Primary guidance will be provided by Prof. Stephen Wu.
Research Project #1: Assessing Disease Progression and Treatment Efficacy for Parkinson’s and Huntington’s Diseases Using Data Analytics on Body Worn Sensors
Parkinson's and Huntington's diseases are characterized by debilitating motion irregularities: such as tremors, unsteady gain, involuntary movements, and lack of coordination. This project seeks to use analytics on data captures from minimally obtrusive sensors worn at multiple points on the body for detecting and classifying motion irregularities, for quantifying the durations of such symptoms, and for characterizing the efficacy of medication in mitigating these symptoms.
Research Project #2: Deep Learning and Data Analytics for Ophthalmic Diagnosis
Common systemic diseases, such as diabetes and hypertension, affect the body's vasculature. These vascular changes can be visualized and assessed using fundus photography (FP) and wide-field fluorescein angiography (FA), a process that involves injecting dye and taking images of it passing through the retinal blood vessels. In this project, we aim to develop an automated computer-aided method for retinal image analysis and ophthalmic diagnosis. We focus on applying deep learning techniques to detect retinal vessels in FP and FA images. We also analyze clinical data for to assess disease progression and treatment and to assist physicians.
Research Project #3: The Full-Spectrum Mobile Experience: UR Color Barcodes
UR researchers have recently invented a color version of the seemingly ubiquitous mobile barcodes that allow us to layer three independent pieces of information within each barcode by using the "spectral diversity" afforded by color printing and capture. In this project, we are looking to develop an easy to use mobile application and to test how the capability for layering can be advantageously used for student activity flyers and other events around campus.
Research Project #4: Adaptive Color Visualization for Color Deficient Observers on Android Smartphones
Around 7-10% of the male population in North America has some form of color deficiency. These viewers often find it difficult to tell the difference between certain colors that appear clearly different to observers with normal color vision. The color deficiency is particularly problematic when it causes a loss of discriminability of different objects or when a color deficient individual must engage in a conversation involving standard color terminology designed for color normal viewers. As increasingly popular personalized imaging devices, smart phones can be used as tools to help color-deficient users overcome their deficiency. As a participant in this research you will help develop and deploy techniques for improved color visualization on Android smartphones by specifically exploiting the adaptivity and personalization that these devices offer.
Research Project #4: Noncoding RNA Gene Search: Unlock the Hidden Information in Genomes
With the wide spread availability of high throughput sequencing technology, vast datasets of genomes are now available to researchers for exploration. Conventional protein coding genes can be located within these large genome data sets with relative ease using BLAST and other alignment tools. Noncoding RNAs (ncRNAs) that serve a direct functional role instead of providing a recipe for protein synthesis, however, present a challenge for genomic analysis. Across species ncRNAs are conserved in secondary structure rather than in sequence and they are therefore not discovered by common sequence alignment based search tools. With the discovery of an increasing number of ncRNAs it is clear that they represent the next frontier in advancing our understanding of the genomes. As a participant in this research, you will develop and evaluate new computational methods for identifying ncRNAs.
Research Project #6: Principled Machine Learning Methods for Multiple Sequence Alignment
The alignment of sequence data is a fundamental task in analyzing genomic data, which shares several commonalities with string comparison. For aligning two sequences, hidden Markov models have emerged as the model of choice for aligning two sequences. For more than two sequences, several heuristics have been proposed that operate on sequences in a pair wise fashion. In this project, we seek an alternate more principled framework for multiple sequence alignment using principled machine learning methods. As a participant in this research, you will be involved in developing, implementing, and evaluating new techniques for multiple sequence alignment.
Research Project #1:Terahertz Photonics
The field of THz science and technology is still in its infancy, but has already gained a very large international interest due to its numerous applications ranging from ultrahigh speed communication systems to medical imaging and diagnostics, industrial quality control, and security screening. In conventional terms, we can talk about the “THz gap,” i.e., a region of the electromagnetic radiation spectrum where it is very difficult to successfully operate either “classical” electronic or photonic devices. For even the fastest FET-type transistor structures, the THz frequency of operation is extremely high, while for optics the THz radiation wavelength is far too long, since the energy of THz quanta is much smaller than the thermal energy at room temperature.
There is an opening for summer research for TWO students. They will be involved in the state-of-the-art research aimed towards generation, propagation, and, subsequent, time-resolved detection of sub-picosecond (THz-bandwidth) electrical transients. Our THz spectroscopy system has an operational frequency bandwidth extending 3.5 THz and is very well suited for ultrafast, noninvasive characterization of various novel electronic materials and nanodevices. As an example, there are our studies of nonequilibrium carrier dynamics in isolated graphene nano-flakes, imbedded in a polymer medium and forming a nanocomposite. We are especially interested in students with interest in optics and electromagnetism, and with a working knowledge of MATLAB.
Research Project #2: Intrinsic Nanodevices
Intrinsic nanostructures exhibit unique physical properties that are a direct consequence of their nano-scale parameters and the quantum mode of operation, in contrast to scaled down versions of well-known conventional devices. One of the best-known examples is a nonlinear, asymmetric nano-channel diode (ANCD), also called self-switching diode (SSD), invented in 2004 by Song et al. Unlike conventional, e.g., p-n diodes, the ANCD performance is not based on a vertical structure and the junction barrier, instead, it produces diode-like characteristics through nonlinear carrier transport in depletion-controlled nano-channel. The ANCD planar geometry allows for a flexible design and easy integration as multi-element sensors. Indeed, ANCDs have been demonstrated to be viable THz detectors and, based on Monte Carlo simulations, are expected to be efficient THz generators.
There is an opening for summer research for ONE student. She or he will be involved in the state-of-the-art research aimed towards testing and, subsequent, analysis and numerical simulation of operation of ANCDs. Our laser-based photoresponse system has a sub-picosecond time resolution; thus, it is ideally suited for ultrafast, noninvasive characterization of various novel electronic materials and nanodevices. As an example, there are our integrated, “experiment-on-chip” studies of nonequilibrium carrier transport in semiconductor structures, as well as ANCDs photoresponse in a single-optical–photon limit. We are especially interested in students with interest in nanoscience and with a working knowledge of MATLAB.
Research Project Descriptions
Research Project #1: Language-Guided Adaptive Grasp Control for Assistive Prosthetic Devices
Assistive prosthetic devices that are able to accept and adapt their behaviors from human guidance have the promise to help people with tasks of daily living with limited computational resources. Previous work in our laboratory studied the problem of developing a wearable prosthetic device that uses multiple sensing modalities (surface EMG, language, and vision) to train and adapt the grasping behavior based on the type of object the prosthetic is interacting with. A limitation of this approach was that the strength of the grasp and the diversity of grasp shapes were fixed across all objects. This project will involve design, implementation, and experimentation of an adaptive grasping controller and prosthetic device that will enable a human operator to create new grasps for a variety of delicate and heavy objects.