Research Projects

This is a list of research projects that may have opportunities for undergraduate students. You can browse all the projects, or view only projects in the following categories:

Computer Engineering and Computer Science

 

Artificial Intelligence: Understanding and extending an implemented reasoning system

Research categories:  Computational/Mathematical, Computer Engineering and Computer Science
School/Dept.: Electrical and Computer Engineering
Professor: Robert Givan
Preferred major(s): Any
Desired experience:   The ideal candidate will be very strong in both mathematical aptitude and computer programming skills. Candidate will learn the LISP programming language as well as formal mathematical logic. Ideal coursework background is ECE 368 and ECE 369. Candidates lacking this background may be considered but will spend more of the summer coming up to speed (perhaps even all of it).
Number of positions: 1

Humans mentally model complex and highly structured environments and effortlessly draw immediate conclusions from what they know about these environments. In this project, we have implemented a system capable of doing exactly these things, though not as well as humans. SURF researchers are invited to learn about this implementation, experiment with it, and propose/implement extensions to the system.

For project background information, go to this website:
https://engineering.purdue.edu/~relation/surf/

 

Artificial vision with neural networks

Research categories:  Computer Engineering and Computer Science
School/Dept.: BME
Professor: Eugenio Culurciello
Preferred major(s): BME, EE, ME
Desired experience:   skilled in programming, lua, torch, C
Number of positions: 1

Design an artificial vision system to learn from a dataset of images and report object class. Also learn actions in movies. ALso learn to play video games with q learning

 

Atomistic Simulations of Gold-Silicon Interface

Research categories:  Aerospace Engineering, Chemical, Civil and Construction, Computational/Mathematical, Computer Engineering and Computer Science, Industrial Engineering, Material Science and Engineering, Mechanical Systems, Nanotechnology, Physical Science
School/Dept.: School of Aeronautics and Astronautics
Professor: Michael Sangid
Desired experience:   Junior standing and ability to develop computer codes.
Number of positions: 1

The size of electronic devices has been decreasing steadily over the years and it is expected to continue that trend, as there is significant interest in the development to microelectronics and nanoelectronics for applications in the biomedical, sensing, data storage and high-performance computing fields, among others. With the increasing miniaturization of electronics, it is important to consider any effects that might happen in the interfaces at the nanometer scale, as the behavior of materials at this length scales may differ markedly from the behavior at the macroscopic scale. This project studies the interactions occurring in the interface between gold and silicon, materials selected due to their excellent properties as conductor and semiconductor, respectively, and their popularity in electronic circuits. The behavior of gold and silicon is expected to differ from the properties observed in the bulk and at larger scales, so it is crucial to analyze and understand the mechanisms of this behavior for the design and manufacture of microelectronic devices utilizing these materials. The research will involve Molecular Dynamics modeling of the gold-silicon interface. Additionally, this project will be complemented by other research opportunities in our lab.

 

CSS Explorer- a cyber-physical virtual reality environment for serious gaming

Research categories:  Computer Engineering and Computer Science
School/Dept.: IE
Professor: Juan Wachs
Preferred major(s): Computer Science, Computer Graphics - Technology, Computer Engineering
Desired experience:   OpenGL, Java, C++. Experience in Computer Graphics is a MUST.
Number of positions: 1

The project consists of developing a virtual environment (using computer graphics) which represents a cyber physical system. The virtual environment is called CSS Explorer, and it is a tool for visualizing cyber-operations. It is composed of aggregated layers displaying information about the status of communication networks between the different components that constitute the networks, such as servers, physical locations of Air Force bases, tasks associated with a mission, etc.

This system will allow us to re-create cyber-attacks. Currently we have game based interfaces that we can use for interaction to respond to a simulated cyber-attack by identifying the threats and re-assigning cyber and
physical resources in order to keep the missions in place.

 

Cloud Computing Research

Research categories:  Computer Engineering and Computer Science
School/Dept.: Electrical and Computer Engineering
Professor: Sanjay Rao
Preferred major(s): Computer Engineering or Computer Science
Desired experience:   Strong programming background. Courses like C-programming, Python programming (ECE 264, ECE 364) would be important (though strong programming background in 1 or 2 languages more important than which specific language). Background in topics like OS, Networking would be a plus.
Number of positions: 1 or 2

While cloud computing offers many attractive advantages, architecting latency sensitive cloud applications remains a challenge. In this project,you will conduct research on cutting-edge cloud technologies, and learn
about and experiment with real cloud platforms such as Amazon EC2.

We also have opportunities on research combining smartphones and the cloud.

 

Constructing Large Scale Representative Social Networks

Research categories:  Bioscience/Biomedical, Computational/Mathematical, Computer Engineering and Computer Science, Educational Research/Social Science, Industrial Engineering, Life Science, Physical Science
School/Dept.: Industrial Engineering
Professor: Mario Ventresca
Desired experience:   At least one student with object-oriented programming, data structures, algorithm analysis, as well as familiarity with at least basic discrete mathematics, statistics and probability (graph theory, distributions, hypothesis testing, regression, etc). Experience using MPI/parallel computation would be highly advantageous, but not necessary. Familiarity with Linux and R would also be useful. A student with a background in epidemiology or applied mathematics would also be very strongly considered.
Number of positions: 1-2

Facebook, Twitter, Orkut and LinkedIn are well known examples of cyber-social networks. However, social networks obviously exist outside of that domain and can represent the connections we make with the people around us. Unlike cyber-world social networks where the connections people make are fully known and observable, real-world networks must be inferred from limited statistical information as well as reasonable assumptions about human behavior. In both instances, a representative social network allows for the study of not only the network topological properties but also diffusive processes acting upon it, such as information spread, influence, disease, etc. There exists a gap in our ability to reconstruct real-world networks from partially observed, noisy and limited data.

This project will aid in developing efficient and scalable algorithms for constructing real-world social network representations at different scales and abstractions (up to global). An extensive literature review of existing capabilities and known social behaviors/mixing patterns will be conducted as part of the project, as will data acquisition and analysis.

 

Continuous Analysis of Many CAMeras (CAM2)

Research categories:  Computer Engineering and Computer Science
School/Dept.: Electrical and Computer Engineering
Professor: Yung-Hsiang Lu
Preferred major(s): Electrical Engineering, Computer Engineering, Computer Science, Mechanical Engineering
Desired experience:   Computer Programming (C, C++, or Python)
Number of positions: 2

The project will retrieve data from publicly available cameras on the Internet and analyze the data. The system currently has more than 30,000 cameras worldwide. Cloud instances are used to analyze the data. The project intends to answer the following questions:

(1) Can cloud computing handle the large amount of streaming data? What parameters would affect the performance?

(2) How to reduce the cost in using cloud instances when analyzing the geographically distributed streaming data?

(3) What useful information can be obtained by analyzing the data from thousands of cameras simultaneously?

 

CyberMech: A Novel Run-Time Substrate for Cyber-Mechanical Systems

Research categories:  Aerospace Engineering, Civil and Construction, Computational/Mathematical, Computer Engineering and Computer Science, Electronics, Mechanical Systems
School/Dept.: Civil Engineering
Professor: Arun Prakash
Preferred major(s): Structural Engineering, Civil, Mechanical, Aerospace, Computer Science, Electrical
Desired experience:   A strong background in the following areas is preferred: Mathematics, Computer Programming, Mechanics, Physics
Number of positions: 1

This project is also a joint collaborative project between myself, Prof. Shirley Dyke from Mech Eng., and faculty from Computer Science at Washington University (St. Louis). In this project, we are developing a computational platform that enables Real-Time Hybrid Simulations (RTHS) of complex structural systems. As opposed to a pure numerical simulation, a hybrid simulation is one where we have a physical specimen of a particular structural component (say a magneto-rheological damper - that is used to control vibrations of structures such as buildings, bridges, automobiles, air-planes, or space structures), that is combined with a numerical model of the entire structure (in real-time) to simulate how this component would behave / control the oscillations of the full structure. This is a handy approach, because it is difficult and expensive to do actual full-scale testing of the component on large scale structures. The challenges associated with this project are first to devise effective coupling mechanisms that allow 'simulating' the physical component (MR damper) as if it were connected in-place within a large structure, and then to develop a computational platform that enables fast, real-time, control and testing of the component combined in different ways with the numerical model of the entire structure.

 

DARwIn Humanoid Robots

Research categories:  Computer Engineering and Computer Science
School/Dept.: Electrical and Computer Engineering
Professor: Yung-Hsiang Lu
Preferred major(s): Electrical Engineering, Computer Engineering, Computer Science, Mechanical Engineering
Desired experience:   Computer Programming (C or Python), Physics, Linear Algebra
Number of positions: 2

In the projects, we aim to answer these research questions:

(1) Can one DARwIn robot use the built-in cameras to recognize objects, use the on-board computer to decide how to lift objects, and then transport the objects? Figure 3 (a) shows an example.

(2) Can two DARwIn robots use the Wifi interface to communicate for lifting and transporting objects. Figure 3 (b) shows an example when two robots transport an object. In this example, the two robots move sideways. There are many other ways for the robots to move. For example, one robot walks forwards and the other robot walks backwards.

(3) How much energy does it take for computation and mechanical movements? What parameters affect the energy consumption? If the on-board computer is too slow, can offloading computation improve the speed in making decisions and thus reduce the energy consumption?

 

Design and evaluation of a data connection pathway and a new user interface between multiple online water quality models

Research categories:  Computational/Mathematical, Computer Engineering and Computer Science, Environmental Science
School/Dept.: ABE
Professor: Lawrence Theller
Preferred major(s): agricultural, natural resources or environmental engineering or software engineering or computer science
Desired experience:   One position is for a student interested in watershed scale management of soil erosion, pesticide loss in runoff, or nutrient management. One position is for a student with classwork or experience with at least *one* of these: HTML 5 development or REST Service - XML schema or Java applications or JavaScript web development or Android development.
Number of positions: 2

This team is investigating the possibility of creating water quality models which are programmed with open-source components (Web processing Services) that can read open format streaming data (Web Feature Services) from various servers, such as the EPA Exchange Network or USGS Realtime gauges. This team will be designing a web application, which is a server-side environmental analysis model. This model will use streaming REST data services as inputs and outputs into a database which then uses open-source software Geoserver to stream the results to the web to be displayed over a backdrop like Google Maps.
Students will work on a team which is testing the computational accuracy of some of our hydrologic modeling tools, and students will participate in refining the user interface of the tools. This will involve minor exposure to some python and javascript although one position is not expected to be doing programming, one intern will be providing user feedback to the (staff and intern) programmers in terms of functionality, ease of use, and accuracy of calculations. For example this intern will run hydrologic models in desktop and online GIS and compare the output results to known and expected outputs, in order to locate problems in calculations or in how the user interface is working. This will be a great position for students interested in watershed scale management of soil erosion, pesticide loss in runoff, or nutrient management.
Depending on background, the second intern will be involved with constructing (programming) either the data handling or computational aspect of an experimental effort to connect two online models to servers streaming water quality data from EPA databases. The project team is conducting research into methods to connect existing water quality models to new types of input data streaming from federal and private servers.

 

Distributed Hybrid Simulation for Dynamical Systems to Natural Hazards

Research categories:  Civil and Construction, Computational/Mathematical, Computer Engineering and Computer Science, Mechanical Systems
School/Dept.: Mechanical Engineering
Professor: Shirley Dyke
Preferred major(s): Mechanical, Civil Engineering; Computer Science or Engineering
Desired experience:   Matlab knowledge is very helpful. Experience with differential equations preferred.
Number of positions: 1

Real-time hybrid simulation (RTHS) is an emerging technique that allows for cost-effective testing of dynamical systems. HS combines physical experimentation with computational simulation to understand how structures and lifelines respond to earthquakes and other natural hazards. RTHS executes this class of test at the actual speed of the earthquake input by using embedded systems with real-time computing capabilities to communicate among experimental and computational resources. In this project we aim to advance the state of the art in RTHS through the use of distributed environments. Both the physical simulation (a building structure) and the computational (controllers and computational models) aspects of the testing require efforts to ensure that a test is conducted in a reliable and accurate manner. Students working on this project will be engaged in learning how to conduct physical experimentation and execute computational models to gain knowledge that will advance the state of RTHS.

 

Mobile Microrobotics

Research categories:  Computer Engineering and Computer Science, Mechanical Systems
School/Dept.: Mechanical Engineering
Professor: David Cappelleri
Preferred major(s): Computer Engineering, Computer Science, Electrical Engineering, Mechanical Engineering or related experience
Desired experience:   Proficiency in C-based language programming, LabVIEW, Image processing (OpenCV), and hardware interfacing (analog or digital I/O). Electronics experience would be helpful.
Number of positions: 1

Mobile microrobots offer unprecedented capabilities for observing and interacting with the world that are not possible with conventional macro-scale systems. A critical issue in the design of mobile microrobots is the generation of wireless power and methods of converting that power into locomotion. We have successfully used externally applied magnetic fields for the power and actuation of individual magnetic mobile microrobots. We have also come up with novel tumbling microrobot designs to overcome the challenge of large surface forces at the micro-scale. In the case of multiple microrobots, all the robots in the workspace will be exposed to identical control signals. Thus, in order to achieve different behaviors from individual robots needed for advanced manufacturing tasks, there must be either significant variation in their design or in the magnetic control signals applied to each microrobot. Therefore, we are working a specialized control substrate for local targeting of the magnetic forces at a fine resolution to be able to independently control multiple microrobots at the same time.

In this project, the SURF student will develop a custom program to control the various types of magnetic microrobots we are working with. Image processing to identify where the microrobot is and then logic to determine what magnetic coil to turn on/off will need to be determined and the appropriate control signals generated. The student should be proficient in C-based language programming, LabVIEW, OpenCV, and hardware interfacing (analog or digital I/O). The SURF student will work with a post-doc on the project.

 

Real-time Lake Michigan buoy research

Research categories:  Civil and Construction, Computer Engineering and Computer Science, Environmental Science, Physical Science
School/Dept.: School of Civil Engineering
Professor: Cary Troy
Preferred major(s): CE, ECE, or ME
Desired experience:   Experience using Matlab is required; student should have some hands-on experience with tools; student should have strong written communication skills; student should like water (boat experience is a plus).
Number of positions: 1

The student will be working on a Lake Michigan buoy that transmits data in real-time on lake water temperatures and meteorological conditions. This buoy is an important source of information for boaters, anglers, and beachgoers along Lake Michigan's southern coast. In particular, the student will work on developing real-time visualizations of lake subsurface water temperatures on a buoy website. In addition, the student will utilize recently-collected buoy data to characterize lake water temperatures, their seasonal variability, and potentially their forecasted changes under various climate change scenarios.

 

Realistic Simulation of Jet Engine Noise using Petascale Computing

Research categories:  Aerospace Engineering, Computational/Mathematical, Computer Engineering and Computer Science
School/Dept.: School of Aeronautics and Astronautics
Professor: Gregory Blaisdell
Preferred major(s): AAE, MATH, CS, ECE, PHYS
Desired experience:   Fourier transforms, computer programming, compressible fluid mechanics (desirable, but not absolutely necessary)
Number of positions: 1

We are currently developing a scalable parallel large eddy simulation code that can realistically simulate high Reynolds number jet flows from complex nozzle geometries. The motivation behind the project is to gain insight into the noise generation mechanisms in a turbulent jet, which is crucial for designing noise reduction solutions such as chevrons. Such high-fidelity simulations generate hundreds of gigabytes of flow-field and acoustics information. As a result, it becomes a significant challenge to extract meaningful information that will improve our understanding of the relationship between the turbulent jet flow and the far-reaching noise it generates. The SURF student will assist us in this effort by developing a set of tools that can help characterize jet noise sources.

A popular model postulates that two distinct noise sources are active in a turbulent jet. One is the large coherent turbulent structures that radiate noise at shallow angles relative to the jet axis, and the second is the fine-scale turbulence that is more dominant in the sideline directions [1,2]. The SURF student will implement several statistical tools that will process the near- and far-field simulation data by computing correlations that are used for examining the source characteristics. These include auto-correlations and cross-correlations of the far-field pressure measurements, as well as correlations between turbulent fluctuations inside the jet and the far-field pressure. These tools will be applied to actual simulation datasets to study how well the noise estimated by our code agrees with the "two-source" model. Furthermore, results from two jet simulations with different boundary conditions will be analyzed to determine the impact on the noise sources.

The SURF candidate is expected to have an interest in Computational Fluid Dynamics (CFD), as well as computer programming. They do not have to have had experience with Fortran, but they must have experience using a computer programming language and be willing to learn Fortran. A strong background in mathematics is also needed. The student will spend some time on the basics of Fortran. This will be followed with a review of the numerical methods used in the code, and the overall structure of the code. The student will then implement the aforementioned capabilities into the code. Finally, the student will apply his or her code to simulation datasets and examine the results.

References:
[1] Tam, C. K. (1995). Supersonic jet noise. Annual Review of Fluid Mechanics, 27(1), 17-43.
[2] Tam, C. K., Viswanathan, K., Ahuja, K. K., & Panda, J. (2008). The sources of jet noise: experimental evidence. Journal of Fluid Mechanics, 615, 253-292.

 

SLEEC: Semantically-enriched libraries for effective exa-scale computation

Research categories:  Aerospace Engineering, Civil and Construction, Computational/Mathematical, Computer Engineering and Computer Science, Electronics, Mechanical Systems
School/Dept.: Civil Engineering
Professor: Arun Prakash
Preferred major(s): Structural Engineering, Civil, Mechanical, Aerospace, Computer Science, Electrical
Desired experience:   A strong background in the following areas is preferred: Mathematics, Computer Programming, Mechanics, Physics
Number of positions: 1

This project is in joint collaboration between myself, faculty in the Electrical and Computer Engineering department at Purdue, and a Computational Research Scientist at Sandia National Labs (Albuquerque NM). What we are doing is trying to improve the performance of library subroutines that are commonly employed to solve problems in solid and fluid mechanics, using finite element methods on very large parallel computers, for instance. Most computational libraries are based on well-formulated mathematical operations, however, when researchers utilize these libraries in their own applications, they are unable to transmit this rich mathematical information to the library and to the underlying hardware. We are devising ways to allow researchers to add/annotate these libraries with useful mathematical information that will allow the computer system to make optimizations on the fly to improve the performance of large computational applications. The challenges associated with this project are first to come up with the right set of mathematical information that can enable such performance improvement, and then to find ways to encode into the libraries in a sufficiently general way so that researchers from different disciplines (solids / fluids) may be able to utilize these libraries to their application programs.

 

Surgical Incision and Suture using the Tauros Robot

Research categories:  Computer Engineering and Computer Science, Mechanical Systems
School/Dept.: Industrial Engineering
Professor: Juan Wachs
Preferred major(s): ECE, ME, Computer Science
Desired experience:   Very good experience programming in C/C++. High GPA. Willing to work with cool robots!
Number of positions: 1

This project is about designing the framework for allowing our existing Taurus surgical robot to perform surgical incision and sutures on a simulator. Currently the system allows control using free hand movements and pedals controls. We are interested in shifting the control to shared (half autonomous, half supervised).
The potential impact of developing the framework for a mobile robot, small, and highly dexterous to support surgery is huge.

The student contribution will be specifically in modeling the surgical tasks, using modeling tools, machining new tools at the micro-scale, and the development of the code for controlling the robot.

 

VACCINE-Visual Analytics for Command, Control, and Interoperability Environments

Research categories:  Computational/Mathematical, Computer Engineering and Computer Science, Innovative Technology/Design
School/Dept.: ECE
Professor: David Ebert
Preferred major(s): Computer Engineering, Computer Science, other Engineering majors with programming experience
Desired experience:   Programming experience in C++, other as described below
Number of positions: 5

We are currently searching for students with strong programming and math backgrounds to work on a variety of projects at the Visual Analytics branch (VACCINE) of the new Department of Homeland Security Center of Excellence in Command, Control and Interoperability. Students will each be assigned individual projects focusing on developing novel data analysis and exploration techniques using interactive techniques. Students should be knowledgeable in C++ on entering the program and will be expected to learn skills in R, OpenGL, and/or a variety of other libraries over the course of the summer.

Current project plans will include iPhone programming for criminal incident report analysis, incorporating local statistics into volume rendering on the GPGPU, and healthcare data analysis. If you have iPhone programming experience or CUDA programming experience or an intense interest to learn either of those, please indicate this in your application form. We also plan to have a project that will assist first responders in accident extrication procedures.

The ideal candidate will have good working knowledge of modern web development technologies, including client-side technologies such as HTML5, SVG, JavaScript, AJAX, and DOM, as well as server side components such as PHP, Tomcat, MySQL, etc. Experience in visualization or computer graphics is a plus. The project will likely be based on the D3 (http://d3js.org/) web-based visualization toolkit; prior experience using D3 or other visualization APIs for the web is particularly welcome.

Of the past undergraduate students that have worked in the center, five of their research projects have led to joint publications in our laboratory and at many of our areas' top venues. Our goal is to continue the center of excellence this summer. Sample projects include visual analytics for law enforcement data, health care data and sports data. Students will be assigned individual projects based on the center's needs which will be determined at a later date. To learn more about the VACCINE Center go to the website provided below.

More information: http://visualanalytics-cci.org

 

nanoHUB Research in Nanoscale Science and Engineering

Research categories:  Computational/Mathematical, Computer Engineering and Computer Science, Electronics, Material Science and Engineering, Nanotechnology, Other
Professor: NCN Faculty
Preferred major(s): Electrical, Computer, Materials, or Mechanical Engineering; Physics; Chemistry
Desired experience:   Programming skills in any language are a plus.
Number of positions: 12

Join the http://nanoHUB.org team and help build the growing set of resources being used in all Top 50 Colleges of Engineering (US News & World Report rankings) and over 240,000 annual users in 172 countries. nanoHUB provides over 260 simulation tools that users run from a web browser in a scientific computing cloud. The Network for Computational Nanotechnology (NCN) operates nanoHUB.

You will work with one of the nanoHUB investigators, including Professors Klimeck, Lundstrom, Alam, Datta, and Strachan and others.

You will learn the Rappture (www.rappture.org) toolkit that makes it quick and easy to develop powerful, interactive, web-based applications. You will work with nanotechnologists to put their applications and supporting information on http://nanoHUB.org. You will test new capabilities in nanoHUB cyberinfrastructure. And you will be part of a National Science Foundation-funded effort that is connecting theory, experiment and computation in a way that makes a difference for the future of nanotechnology and the future of scientific communities. Other undergraduate researchers before you have each been able to literally impact over a thousand nanoHUB users (for an example, see https://nanohub.org/resources/normdist); join their legacy and create something that will help your own skills and will help others.

More information: http://nanoHUB.org