Research Projects for Summer 2021

Duke's Pratt School of Engineering is offering research experience opportunities across each its four academic departments:

Grand Challenge REU participants have the opportunity to conduct research across a large spectrum of interdisciplinary topics broadly organized into five areas: energy, environment, health, national security, and learning—

Five topic areas of Grand Challenges REU program

The following is a list of research projects available during the summer 2021.

COVID UPDATE: This year we are offering a virtual REU program.

Inverse problems and Optimization Modeling

Grand Challenge: Engineering the tools of scientific discovery

Faculty Mentor: Wilkins AquinoProfessor in Mechanical Engineering and Materials Science

The student will work closely with Prof. Wilkins Aquino and members of the IoMechlab (Inverse Problems and Optimization in Mechanics) on one or more projects at the intersection of physics-based computer simulation, optimization, inverse problems, and machine learning.  Current applications include characterization of vascular disease using ultrasound vibrometry, autonomous anomaly detection in nuclear reactors using mobile robots, and modeling and optimization of the human voice, among others.

Network based approaches to understand neurodegenerative diseases

Grand Challenge: Reverse engineer the brain

Faculty Mentor: Alexandra Badea , Associate Professor in Radiology

The goal of this project is to familiarize students with interdisciplinary approaches to understand the brain. The main thrust is on developing computational approaches using graph analysis and visualization of diffusion tensor imaging and derived tracts. We will use high resolution diffusion tensor brain imaging. The goal is to use animal models to help better understand human neurodegenerative diseases, such as Alzheimer's disease. The student will use R, MATLAB, or python programming for graph analysis (or his /her choice of a programming language), and compare several options for visualizing and assessing the quality of reconstructed tractography data. I am interested in supporting the development of computational skills in engineering students, with a particular focus on supporting female students. 

Data-driven Polymer Nanocomposite Design via the Materials Genome Initiative (Microstructure Characterization Tools – Database Pipeline)

Grand Challenge: Engineer the Tools of Scientific Discovery

Faculty Mentor: L. Cate Brinson , Professor in Mechanical Engineering and Materials Science

The Brinson laboratory specializes in both computational and experimental characterization, as well as design of nanostructured polymers and polymer nanocomposites. A major challenge in this area is the nearly infinite design space due to the enormous range of materials selection, surface chemistries, processing methods, etc. The Materials Genome Initiative (MGI) is a large-scale, data-driven project with the goal of predicting properties through a combination of data mining and physics-based modeling in order to reduce the deployment time of new materials compared to the traditional trial-and-error experimental approach. The Brinson Lab has pioneered MGI efforts for the informed design of polymer nanocomposite materials through the collaborative development of an open-source data resource (NanoMine/MaterialsMine) that incorporates a database along with a suite of simulation, analysis, and data mining tools. The database contains repositories of curated literature and lab-generated data formatted for data mining and statistical analysis. The microstructure characterization and reconstruction (MCR) tools extract insights from microstructure images by the form of descriptors. This materials informatics approach illustrates an example of utilizing advancements developed in the era of “Big Data” to solve practical materials science problems. For this project, a student is expected to 1) organize their data collection and curation in a repository, 2) understand the workflow of NanoMine database and MCR tools, 3) work with the team to integrate MCR tools into the curation process, 3) work with the team to upload and make accessible the data on the NanoMine/MaterialsMine platform, and 4) work with the team to develop tools for image refinement with machine learning techniques. An ideal candidate should have strong interest in software development, data science, experience reading scientific literature, robust organizational skills, and be both creative and attentive to detail. Familiarity with polymers and polymer nanocomposites is a plus.

Data-driven Polymer Nanocomposite Design via the Materials Genome Initiative (NanoMine Data Curation)

Grand Challenge: Engineer the Tools of Scientific Discovery

Faculty Mentor: L. Cate Brinson , Professor in Mechanical Engineering and Materials Science

The Brinson laboratory specializes in both computational and experimental characterization, as well as design of nanostructured polymers and polymer nanocomposites. A major challenge in this area is the nearly infinite design space due to the enormous range of materials selection, surface chemistries, processing methods, etc. The Materials Genome Initiative (MGI) is a large-scale, data-driven project with the goal of predicting properties through a combination of data mining and physics-based modeling in order to reduce the deployment time of new materials compared to the traditional trial-and-error experimental approach. The Brinson Lab has pioneered MGI efforts for the informed design of polymer nanocomposite materials through the collaborative development of an open-source data resource (NanoMine/MaterialsMine) that incorporates a database along with a suite of simulation, analysis, and data mining tools. The database contains repositories of curated literature and lab-generated data formatted for data mining and statistical analysis. This materials informatics approach illustrates an example of utilizing advancements developed in the era of “Big Data” to solve practical materials science problems. For this project, a student is expected to 1) organize their data collection and curation in a repository, 2) integrate other tools being developed by the team into the curation process, 3) work with the team to upload and make accessible the data on the NanoMine/MaterialsMine platform. An ideal candidate should have strong interest in data science, experience reading scientific literature, robust organizational skills, and be both creative and attentive to detail. Familiarity with polymers and polymer nanocomposites is a plus.

Machine Learning-Assisted Understanding of Polymer Nanocomposite Composition—Structure Relationship

Grand Challenge: Engineer the Tools of Scientific Discovery

Faculty Mentor: L. Cate Brinson , Professor in Mechanical Engineering and Materials Science

The Brinson laboratory specializes in both computational and experimental characterization, as well as design of nanostructured polymers and polymer nanocomposites (PNCs). A major challenge in this area is the nearly infinite design space due to the enormous range of materials selection, surface chemistries, processing methods. Compositional parameters such as matrix and filler selection, filler surface treatment, and filler volume fraction play a critical role in determining the microstructure of PNCs. Molecular modeling techniques such as physics-based Molecular Dynamics (MD) simulations can reveal the microstructure of PNCs by simulating an annealing process analogous to the processing steps in experiments. For this project, a student will perform MD simulations of PNCs with a range of compositional parameters, characterize the dispersion state of nanofillers from MD simulations, as well as interpret the composition—structure relationship using machine learning methods. An ideal candidate should have interest in data science and molecular modeling, have basic knowledge of polymers and experience with programming languages such as Python. Interest in polymer nanocomposites and knowledge of MD modeling is a plus. REU student will be mentored closely to learn fundamentals of MD, machine learning approaches, and structure-property relationships.

Autonomous systems 

Grand Challenge: Restore and Improve Urban Infrastructure

Faculty Mentor: Mary Cummings , Professor in the Department of Electrical and Computer Engineering

Research in the Humans and Autonomy Lab (HAL) focuses on the multifaceted interactions of human and computer decision-making in complex sociotechnical systems with embedded autonomy. Given the explosion of autonomous technology in aviation, medicine, and even in everyday mundane environments like driving, the need for humans as supervisors of and collaborators in complex autonomous control systems has replaced the need for humans in direct manual control. Students who work on projects in HAL will learn to employ human-systems engineering principles to autonomous system modeling, design, and evaluation, and identify ways in which humans and computers can leverage the strengths of the other in an autonomous system to achieve superior decisions together. 

Predictive modeling for ICU admissions after high-risk surgeries 

Grand Challenge: Restore and Improve Urban Infrastructure

Faculty Mentor: Mary Cummings , Professor in the Department of Electrical and Computer Engineering

Some surgical procedures carry high risk, especially for infants, so it would be very helpful if doctors had a predictive model that could indicate during the course of a complex surgery whether a patient should be admitted immediately to an intensive care unit. This effort will attempt to develop such a model using machine learning techniques for craniosynostosis cases. This project will allow students to not only develop state-of-the-art predictive models, but also learn how to work with experts in the formulation and interpretation of model outcomes. An introductory machine learning course is a perquisite for this project. 

Materials Informatics and Visualizations

Grand Challenge: Engineering the tools of scientific discovery

Faculty Mentor: Stefano Curtarolo , Professor in the Department of Mechanical Engineering and Materials Science

AFLOW is a software framework for materials discovery, powering aflow.org and the largest database for inorganic materials (3M+ entries, 500M+ properties). Features are employed for thermodynamic descriptor development and machine learning analyses. Such models have fueled the synthesis of new rare-earth-free magnets and high-hardness carbide systems, both materials being the first of their kind to be discovered by prediction.

Several projects are available: property-workflow development, machine learning and data-mining, creation of online applications (interactive analyses and visualizations).

The ideal candidate must: love coding (C++, Python, HTML, JavaScript, CSS), have passion for materials science (thermodynamics, physics, and chemistry) and data-mining (machine learning), be self-motivated and a critical thinker.

 img

For additional information, please visit http:// aflow.org  

Enhancing Virtual Reality: Intelligent Augmented Reality

Grand Challenge: Enhance virtual reality

Faculty Mentor: Maria Gorlatova , Assistant Professor of Electrical and Computer Engineering

Augmented reality (AR) is a rapidly developing technology area with potential for transforming many daily human experiences. While promising, current AR systems are somewhat limited in their capabilities, in particular in multi-user experiences, high energy consumption, and general lack of robustness and adaptability. The goal of this project is to obtain in-depth experimental understanding of the limitations of current augmented reality experiences, and to establish how these limitations can be addressed.

Students involved in this work will experiment with develop and experiment with different augmented reality applications, experiences, and platforms, in order to understand the systems and network loads of different operations, key drivers of immersive user experiences, and the potential of edge and cloud computing platforms to address the discovered limitations. The project involves both the development of Unity-based holographic experiences, and real-world holographic deployments with Google ARCore mobile device platforms and Magic Leap One headsets.

The project is best suited for students who have an experimental mindset and who enjoy obtaining in-depth understanding of system performance, physical phenomena, and human behavior. Relevant technical preparation includes: General software development skills Coursework in communications and/or networking preferred but not required Coursework in machine learning preferred but not required.

For additional information, please visit https://maria.gorlatova.com/current-research/  

Optimized Parameters of Electrical Stimulation of the Nervous System

Grand Challenge: Re-engineering the brain

Faculty Mentor: Warren Grill , Professor in Biomedical Engineering

The Grill lab specializes in analysis and design of devices that use electrical stimulation of the nervous system to restore function to individuals with neurological impairment. This project focuses on analysis and design of the parameters of stimulation, for example to treat the symptoms of Parkinson’s disease or to treat chronic pain. The project will make use of computer-based modeling. Students working in this research project will develop an understanding of the principles of electrical stimulation and the application of computer modeling to study electrical stimulation. Student with interest in neuroscience and medical devices are ideal candidates for this project. The students will work directly with a PhD student and / or Research Scientist and have the opportunity to interact with a diverse group of students and staff working across a broad array of neurodevices. Previous experience with writing computer code is extremely helpful.

Bacterial electrophysiology

Grand Challenge: Engineer better medicines

Faculty Mentor: Christine Payne , Associate Professor in Mechanical Engineering and Materials Science

All cells have a resting membrane potential driven by an ion gradient. We will use this parameter to control bacterial growth and biofilm formation. The use of electricity, instead of drugs, provides a new method to control these parameters. Students will work with a team to culture bacteria, etch gold electrodes, and measure bacterial growth. While no specific skills are required; students must be curious, independent, and hard working.

For more information, please visit http://payne.pratt.duke.edu/  

Determining the Role of Myeloid Cells in the Brain Tumor Micorenvironment

Grand Challenge: Re-engineering the brain

Faculty Mentor: John Sampson , Professor of Neurosurgery in the School of Medicine and

Brain Tumors are an aggressive tumor, and improvements in standard of care have been limited. Other cancers have recently benefited from the development of immunotherapies, but brain tumors appear to be resistant to current immunotherapies (checkpoint blockade, CAR-T). This resistance is partially because the cells that typically clear tumors (CD8 effector T-cells), have difficulty entering the tumors and once they get there they quickly lose the ability to clear tumor cells because they are exhausted. A group of immune cells that don't have difficulty entering the tumors and contribute to exhaustion of T cells are called Tumor Associated Myeloid Cells (TAMs). This project will focus on better understanding the tumor promoting functions of TAMs, the mechanisms driving these functions, and ways to therapeutically target these functions.

For additional information, please visit https://bme.duke.edu/faculty/john-sampson/