PCE News

PCEC Students Share Research Findings at Student Scholars Day

April 13, 2022

PCEC Students Share Research Findings at Student Scholars Day

Several Padnos College of Engineering & Computing (PCEC) students shared the results of their faculty-mentored research projects at the 26th Annual Grand Valley State University Student Scholars Day on April 13, 2022. Student presenters and their project abstracts are listed below.

 

Improving Engineering Design Education by Fostering a Better Understanding of Electromechanical Elements

This project will create boards that explain common electromechanical elements used in common products and machinery. The boards will be two feet by four feet and hang on the walls of the Innovation and Design Center. Mounted on the board will be a physical example of the electromechanical element. When possible, the electromechanical element will be displayed disassembled or with covers cut away to reveal the inner workings. All parts will be labeled with their name and function. A technical description of how the electromechanical element works will be provided.  

Presenter: Kevin Lott 

Mentors: John Farris, Engineering & Christopher Pung, Engineering

 

A New Space-Filling Curve

Space-filling curves have many applications in scientific computing. Most space-filling curves are displayed using cartesian grids and they work in conjunction with traditional quadtrees. We have proposed a new kind of curve, the LTD curve. We have written a grammar that can generate different levels of recursion of the curve. We have chosen to use barycentric coordinates instead of traditional cartesian coordinates, and we have devised a new flavor of quadtree designed for triangular grids. Previous work on this subject has been surveyed by Michael Bader in his book Space-Filling Curves: An Introduction with Applications in Scientific Computing.

Presenter: Elise Dettling

Mentor: Christian Trefftz, Computing

 

Detecting Face Mask Usage Trends in Social Media with Machine Learning

The use of face masks to prevent disease spread among the general population has become widespread due to the COVID-19 pandemic. The ability to accurately detect and monitor the trends in face mask usage is crucial to understanding and predicting hotspot areas for both current and future pandemics. This work investigates the detection of face masks in social media images using deep learning. This project creates a social-media-based face mask image dataset that reflects the scale needed for deep learning and the diversity (mask types, positions of people, and ethnicity) of real life. We have gathered approximately 120k images containing people from different cities. Mechanical Turk is used to label the images based on the presence of a face mask. Using this dataset, we train a CNN model to detect the presence of face masks in social media images and compare the results to existing datasets. We then deploy our model to detect trends in face mask usage in Los Angeles over time.

Presenter: Seth Ockerman

Mentor: Erin Carrier, Computing

 

Exploring Fast Emulators for Long-running Mars Simulation Code

The NASA Ames Global Climate Model (GCM) software has been in steady use at NASA for decades and was recently released to the public. This model simulates the complex interactions of different weather cycles that exist on Mars, namely the Dust Cycle, the CO2 Cycle, and the Water cycle. It is used by NASA to help understand their empirically observed data through the use of sensitivity studes. However, these sensitivity studies are computationally taxing, requiring weeks to run.  To address this issue, we have developed a surrogate model using Gaussian processes that can emulate the output of this model with relatively small amounts of data in a reduced amount of time (on the order of minutes).  We demonstrate the effectiveness of our emulator using backward error analysis.

Presenter: Marc Tunnell

Mentors: Nathan Bowman, Computing & Erin Carrier, Computing

 

Finding Malicious Malware using Twitter Scanner

Zoom has become an increasingly popular application. This has led to an increase in cyber criminal activity. Cybercriminals have been utilizing social engineering tactics to persuade users to download compromising software. Their agenda is to gain access to sensitive information and exploit your system. I will be utilizing a Python bot to find malicious malware posted on Twitter under the domain name of Zoom.  This bot can scan and identify potential threats, with the idea being to bring awareness to phishing scams and help mitigate dangerous Zoom downloads on Twitter.

Presenter: Morgan Hamlin

Mentor: Erik Fredericks, Computing

 

Identifying Tradeoffs in Benefits of Interaction and Visualization Techniques for Analytic Tasks

In visual analytics research, new visualization or interaction techniques are typically developed to support specific analytic tasks, such as clustering observations of high-dimensional data. User studies on these new techniques seek to demonstrate their effectiveness on specific tasks. Developers of visual analytics systems then seek to combine these techniques to support more tasks, such as characterizing distributions and clustering observations. In this example, the effectiveness of a clustering technique on characterizing distribution tasks was not evaluated and may be detrimental to this task. Therefore, a different technique for this task may be preferred. As such, we promote a broader perspective when evaluating the effectiveness of a new technique so that the tradeoffs in the benefits of using such techniques across a wide range of analytic tasks can be understood. This presentation demonstrates how to complete such an evaluation on the techniques employed by SIRIUS.

Presenters: Calob Horton & Shweta Dnyaneshwar Terkar 

Mentor: Michelle Dowling, Computing

 

Mitigating the Impact of Run-Time Software Engineering

Cyber-physical systems (CPSs) experience uncertainty in all phases of the software development life cycle, generally stemming from real-world concerns such as minimizing the impact of environmental interactions and ensuring the safety of human participants. These systems are often described as safety-critical and therefore must continuously exhibit correct behaviors. Run-time testing can provide assurance that system requirements are satisfied even in the face of uncertainty. Previously we introduced GreenRoom, a proof-of-concept experimental CPS testbed for monitoring the impact of run-time testing on power consumption of a target CPS. This project extends GreenRoom’s capabilities to enable deeper studies into monitoring power consumption by including additional mechanical components. We also intend to investigate how search-based testing techniques can extend CPS abilities, and to monitor their impact on power consumption at run time.

Presenter: Abigail Diller 

Mentor: Erik Fredericks, Computing

 

View the PCEC Student Scholars Day photo album.

Visit the Student Scholars Day website.

Share this news story

View More PCE News



Page last modified April 13, 2022