Research Projects

Projects are posted below; new projects will continue to be posted through February. To learn more about the type of research conducted by undergraduates, view the 2018 Research Symposium Abstracts.

2019 projects will continue to be posted through January!

This is a list of research projects that may have opportunities for undergraduate students. Please note that it is not a complete list of every SURF project. Undergraduates will discover other projects when talking directly to Purdue faculty.

You can browse all the projects on the list or view only projects in the following categories:

Computer Engineering and Computer Science

 

Active Learning: Choosing the Right Data for Machine Learning

Research categories:  Computer Engineering and Computer Science
School/Dept.: Electrical and Computer Engineering
Professor: Yung-Hsiang Lu
Preferred major(s): Computer Engineering, Computer Science, Electrical Engineering, Computer and Information Technology
Desired experience:   Computer Programming

Machine learning can be classified into different categories. One is called supervised learning: each piece of data is associated with a correct answer (also called label). Since machine learning is not perfect, labeling usually needs human efforts and can be very expensive. Another type of learning is unsupervised learning: there is no correct answer and this is frequently used in clustering data into groups. Active learning is somewhat in between. Unsupervised learning is used to cluster data and identify the data that is distinct and should be labeled.

This project will use public datasets of images (or videos) as the foundation for training machine models (supervised learning). Then, new data is clustered to discover which should be labeled. This is part of the CAM2 (Continuous Analysis of Many CAMeras). CAM2 discovers, retrieves, and analyzes vast amounts of real-time data from worldwide network cameras.

More information: https://www.cam2project.net/

 

Applications of Deep Reinforcement Learning

Research categories:  Computer Engineering and Computer Science
School/Dept.: IE
Professor: Vaneet Aggarwal
Preferred major(s): CS, EE, Math, Stats

Reinforcement learning refers to goal-oriented algorithms, which learn how to attain a complex objective (goal) or maximize along a particular dimension over many steps; for example, maximize the points won in a game over many moves. Our group is involved in coming up with fundamental algorithms for many aspects in reinforcement learning, and applying them to problems in social networks and transportation.

 

Building Computer Systems Software for AI and IoT

Research categories:  Computer Engineering and Computer Science
School/Dept.: ECE
Professor: Felix Lin
Preferred major(s): ECE or CS
Desired experience:   Enthusiasm in programming, exploring, and building.

While the mobile computers are still flourishing, we are quickly embracing a variety of new computing paradigms -- wearables, IoT, and VR headsets, just to name a few. They challenge the way we design and build computer software today. In this project, you will be involved in redefining systems software to make computers smarter, faster and cooler. This project will give you a lot of fun in hacking Linux, frameworks, and various modern hardware.

Underrepresented minority students are particularly encouraged to apply.

More information: http://xsel.rocks

 

Building Software for Environmental Modeling

Research categories:  Agricultural, Computer Engineering and Computer Science, Environmental Science, Other
School/Dept.: Agricultural and Biological Engineering
Professor: Dharmendra Saraswat
Preferred major(s): Agricultural Engineering , Civil Engineering, Computer Science or related disciplines
Desired experience:   Programming skills in any language with some experience in frontend and backend web development is desired.

Agricultural and Biological Engineering Department has contributed several tools for environmental modeling community. It is a challenge to review and understand old codes with minimum documentation. This project involves modernizing an environmental modeling software written primarily in Perl. In this project, the SURF student will first assess the current application, create a plan for the new iteration in collaboration with the project supervisor, get a head start on developing the new application and document the process. The SURF student will work with a staff programmer.

 

Data Visualization and Analysis for IoT Based Smart Irrigation System

Research categories:  Agricultural, Civil and Construction, Computer Engineering and Computer Science, Environmental Science, Other
School/Dept.: Agricultural and Biological Engineering
Professor: Dharmendra Saraswat
Preferred major(s): Agricultural Engineering, Civil Engineering, Environmental Engineering, Computer Science or related disciplines
Desired experience:   Programming skills in any language with some experience in statistics is desired.

It is reported that currently almost 33 percent of the global population is affected by water scarcity and by 2030, this figure is expected to climb up to almost 50 percent. Around 60 percent of the water used for irrigation is wasted, either due to evapotranspiration, land runoff, or simply inefficient, primitive irrigation application methods. This realization has brought attention to smart irrigation – powered by the internet of things (IoTs) – that can be a better way of managing water stress on a global basis. In this project, the SURF student will customize commercially available software to analyze and visualize data, perform calculations/combine new data, run time-based calculations, plot functions for visual understanding and perform sophisticated analysis by combining data from several field nodes. The SURF student will work with Project Supervisor and a staff programmer.

 

Geodesic convolution with various applications in 3D data analysis

Research categories:  Computational/Mathematical, Computer Engineering and Computer Science, Mechanical Engineering
School/Dept.: Mechanical Engineering
Professor: Min Liu
Preferred major(s): ME, ECE, CS
Desired experience:   python, c++ code, experience with convolutional neural networks

The scope this project is to explore the mechanics of geodesic convolution (in contrast to the standard Euclidean space convolution) for deep neural networks. The objective is to research for a more efficient, robust and shape-aware filter to support various applications for 3D vision data analysis, E.g. Autonomous CAR, robot navigation, and Augmented realities.

 

Human Body Communication

Research categories:  Bioscience/Biomedical, Computer Engineering and Computer Science, Electronics
School/Dept.: ECE
Professor: Shreyas Sen
Preferred major(s): ECE, BME

The student will work on theory and device design related to using the human body as a communication medium to improve Healthcare and HCI.

 

Human Factors Considerations: Older Adults and Autonomous Vehicle Systems

Research categories:  Computer Engineering and Computer Science, Industrial Engineering, Innovative Technology/Design
School/Dept.: Industrial Engineering
Professor: Brandon Pitts
Preferred major(s): Industrial Engineering
Desired experience:   Human Factors, Matlab, Transportation, some experience in statistics, some computer programming experience (in any language)

Automobiles are becoming increasingly autonomous. At the same time, the demographics of drivers using these advanced vehicles is changing. In particular, adults aged 65 years and older are the fastest growing age group worldwide and are expected to benefit from vehicle automation. However, age-related perceptual and cognitive difficulties may limit the extent to which these systems are useful for individuals in this age category. The goal of this project is, therefore, to quantify interactions between (older adult) drivers and autonomous driving systems in order to develop approaches that enhance roadway safety for various aging populations.

The SURF student will assist with collecting and analyzing data from human-subject experiments (using a laboratory driving simulator) and with writing any project publication. In addition, the student will meet regularly with faculty and graduate mentors to communicate his/her progress.

 

Illumination of Damage through Microtomography

Research categories:  Aerospace Engineering, Computer Engineering and Computer Science, Industrial Engineering, Material Science and Engineering, Mechanical Engineering
School/Dept.: Aeronautics and Astronautics
Professor: Michael Sangid
Preferred major(s): AAE, ME, MSE, EE, CSE, or IE
Desired experience:   Students are expected to work with Image Processing and Visualization tools, as well as Matlab.

Damage in structural materials is often difficult to quantify, instead we rely on large scale component level testing and curve fitting. With the advent of advanced microtomography, we have the ability to identify damage inside the bulk of the material, in which the samples are subjected to mechanical loading. Thus, in this project, microtomography scans will be reconstructed and the damage in the form of voids or cracks will be characterized and quantified in several material systems (including carbon fiber reinforced composites and Ti-6Al-4V produced via additive manufacturing). The interaction of damage with microstructural features will be assessed, in order to achieve a physics-based understanding of material failure.

 

Processing of innovative satellite remote sensing data for ocean and snow remote sensing

Research categories:  Aerospace Engineering, Computer Engineering and Computer Science, Electronics, Environmental Science, Physical Science
School/Dept.: AAE
Professor: James Garrison
Preferred major(s): ECE, AAE, Physics, EAPS
Desired experience:   Good programming skills, signal processing (ECE 301 or AAE301). Experience with software defined radio (USRP) will be a plus.

Reflectometry is a new approach to Earth remote sensing in microwave frequencies, using reflections of Global Navigation Satellite System (GNSS, e.g. GPS, Galileo, etc ...) signals from land and ocean surfaces as illumination source in a bistatic radar configuration. Through observing measurable changes in the properties of these signals, various features of the reflecting surfaces can be inferred.

Ocean surface winds is the most developed application for GNSS-Reflectometry (GNSS-R), with the launch of the CYGNSS constellation by NASA in 2016. CYGNSS data has been collected during the 2017 and 2018 Hurricane seasons, showing some capability for wind field measurements at a high spatial resolution. New models and algorithms are required, however, to optimally process these data and extract wind vectors with high sensitivity, especially at the higher wind speeds present in hurricanes. Development of these new models and algorithms requires the collection of high-quality data under carefully controlled conditions along with in situ training data provided by independent sources. With this goal in mind, Purdue has developed a wideband GNSS-R signal recorder which will be flown on the P-3 “Hurricane Hunter” aircraft operated by NOAA. This aircraft is capable of operating in extremely high winds and penetrating the Hurricane eye wall, in order to collect data inside developing tropical cyclones. GNSS-R data collected in this experiment will be compared with wind speed observations from other instruments on the P-3 aircraft, other satellite data, and model results. These comparisons will be used to develop and improved model for the extraction of ocean winds from CYGNSS and future satellite missions.

Snow Water Equivalent (SWE) is a representation of the total water stored in the snow pack. This is an important climate variable for the prediction of fresh water supplies as well as applications such as hydroelectric power. A new application of GNSS-R is measuring SWE as a change in phase of the reflected signal, a result of the slower propagation of the signal through the snow layer. Spaceborne measurements of SWE using GNSS-R have never been conducted. Special collections of CYGNSS data were conducted this year, in which raw signals (no on-board processing or compression) were collected in arcs spanning snow-covered regions in the Himalayan mountains.

SURF projects are proposed to support these two research goals for CYGNSS data. Both will involve extensive programming and data processing, using a “software defined radio” method that essentially implements all signal processing in software to operate on the full-spectrum of the recorded signal.

Applicants should have very strong programming skills, some knowledge of basic signal processing.

 

Programming 3D and environmental data acquisition into iFly -- a mobile iOS app

Research categories:  Computer Engineering and Computer Science, Life Science
School/Dept.: Entomology
Professor: Trevor Stamper
Preferred major(s): Computer science or engineering, or biological sciences
Desired experience:   Must have programming knowledge in Swift programming language. Mobile device iOS programming experience is highly desired.

The student researcher will be programming in Swift language on the iFly project to allow environmental sensor systems and 3D sensing systems to input data directly into the app. Student researcher will also be improving other functions of the software to build a better user experience.

 

SMART (Social Media Analytics Response Toolkit)

Research categories:  Computational/Mathematical, Computer Engineering and Computer Science, Innovative Technology/Design
School/Dept.: Electrical and Computer Engineering
Professor: David Ebert
Preferred major(s): Computer Science, Electrical and Computer Engineering
Desired experience:   For this project, the ideal candidate will have good working knowledge of some of the modern web development technologies, including client-side technologies such as HTML5, SVG, JavaScript, AJAX, and DOM, and D3 as well as server-side components such as PHP, Tomcat, MySQL, etc. Experience in web services development and web based visualization APIs is a plus. Students should have a GPA of 3 or higher.

This visual analytics application provides interactive (Twitter) social media analysis and visualization capabilities through topic extraction, combination of filters, cluster analysis and stream categorization. Analysts can also create custom classifiers to extract social media messages relevant to specific events or topics. Many first responder groups in the U.S. use this platform.

 

Sensing the Human Factors in Laparoscopic and Robotic Surgery

Research categories:  Bioscience/Biomedical, Computer Engineering and Computer Science, Industrial Engineering, Mechanical Systems
School/Dept.: Industrial Engineering
Professor: Denny Yu
Preferred major(s): Industrial Engineering, other
Desired experience:   Human Factors, Matlab, Machine Learning, Healthcare, Medical Device Design

Work-related musculoskeletal disorders (MSDs) among surgeons are becoming more common. The purpose of this project is to use sensors to measure ergonomic risks and assess interventions to surgeons during laparoscopic and robotic surgery. This work will leverage sensing technology (e.g., motion tracking, pressure map, electromyography) to monitor surgeons’ ergonomics to ultimately develop recommendations on minimizing MSDs and how to better design an operating room.

The SURF student will participate in data collection in the operating room at Indiana University School of Medicine, data analysis and interpretation, and write his/her results for a journal publication. The student will regularly communicate his/her progress and results with faculty, graduate mentors, and surgeon collaborators.

 

Smart Manufacturing using IoT and Machine Learning

Research categories:  Computer Engineering and Computer Science, Innovative Technology/Design, Mechanical Engineering
School/Dept.: Mechanical Engineering
Professor: Martin Jun
Preferred major(s): Mechanical Engineering, Computer Engineering, or Computer Science
Desired experience:   Virtual reality programming, mechatronics, CAD design and programming for graphics, signal processing and data analysis, machining, etc.

Autonomous operation and decision making during manufacturing processes and production are important. Using IoT technologies, machine-to-machine, machine-to-human communication and data generation are achieved and machine learning algorithms are used for data analysis and decision making. The student will work on virtual reality (VR) based visualization of data achieved from IoT devices connected to CNC machine and robots and analyze data using machine learning.

 

The Arequipa Nexus Sustainable Viticulture

Research categories:  Agricultural, Computational/Mathematical, Computer Engineering and Computer Science, Environmental Science, Innovative Technology/Design
School/Dept.: Electrical and Computer Engineering
Professor: David Ebert
Preferred major(s): Flexible: Computer Science, Food Science, Agronomy, Environmental Science, GIS, Electrical and Computer Engineering
Desired experience:   We are looking for applicants with a strong background in either of the following: GIS (Geographic Information Systems), food sciences, agronomy (soil oriented), web development or python programming (e.g. HTML/JavaScript, Leaflet, D3). Students should have a GPA of 3 or higher. Applicants with Spanish fluency are encouraged to apply.

The Universidad Nacional de San Agustín (UNSA) in Arequipa, Peru and Purdue through Discovery Park’s Center for the Environment (C4E) have partnered to create a new research, education and innovation institute to work together on key challenges for a sustainable future for the citizens of Arequipa. The Nexus Institute applies collaborative, data-driven, interdisciplinary science, technology and innovation to help chart a new course toward a sustainable future. Our lab works with key stakeholder groups to develop data, provide (winery and vineyard farm) guidelines, simulation models, and decision support tools for vineyard management through state-of-the-art data sets, GIS and remote sensing, and environmental decision tools. We are also developing a system to provide farmers with more accurate information than previously possible, helping growers to optimize crop yields and minimize use of water and other resources. The system will be first tested in Peru to create precision agriculture-based viticulture test-beds.