2024 Design and Research Conference

Home › 2024 College of Engineering and Science Design and Research Conference › Cyber Engineering Senior Projects

Cyber Engineering Senior Projects

Integrated Engineering and Science Building 214.

1:00 p.m.

V.A.R.P. (Vulnerability Assessment and Remote Patching) for Kubernetes

Team Members:  Ethan Dunning, Kirkland Grace, Johnathon Sheffield, Jacob Smart, Nicholas Winkelmann,

Advisor: Dr. Miguel Gates

V.A.R.P. is the blending of security analysis and patching software to be used in cloud infrastructure, specifically within Kubernetes. All functionality will be readily understandable and usable for customers with minimal security knowledge. The security analysis will begin with a full-scope scan of all resources in the cluster. As part of the scan, all commands being run by the program will be logged into a cloud storage solution so that the program can keep track of changes and problems. This information will also be used as input into AI functionality to generate reports of security concerns. Then customers will be prompted to begin patching security concerns. This will be done by Kubernetes command scripts that will harden the cluster.
1:30 p.m.

W.A.V. (Web-based AI Voice Recognition and Authentication)

Team Members:  Wilbert Collins, Logan Johnston, Garret Miculek, Dustin Smith

Advisor: Dr. Miguel Gates

W.A.V. (Web-based AI Voice Recognition and Authentication) is a two-factor authentication system utilizing personalized artificial intelligence to authenticate logins via vocal recognition. The project utilizes a Python-based artificial intelligence system trained on user-generated data to recognize and authenticate the user’s voice and a secure access controls system to handle logins and account creation. The system is designed to be integrated as part of a larger website and managed by administrators dynamically, without extensive knowledge of artificial intelligence or access control security.
2:00 p.m.

Local Home Assistant

Team Members:  Lucas Feazel, Clay Hopkins, John Peterson, John Waskom

Advisor: Dr. Miguel Gates

PRISM (Personalized, Responsive, and Intelligent Support Module) is an alternative solution to other at-home voice assistants (i.e., Amazon Echo, Google Assistant, etc.) that aims to fully respect the user’s privacy and provide a physical medium for automation and control. PRISM’s concept is to provide a smart IoT-powered home assistant without relying on third parties from the cloud to provide services. This is achieved by hosting all the necessary services and programs locally within PRISM’s hardware. PRISM consists of Raspberry Pi’s being the central units with smaller base units, or “modules”, allowing for specific modular sensing capabilities desired by the user. Along with the hardware, PRISM will have a GUI on a small screen, a web server hosting a UI for configuration, and a database temporarily storing necessary data between all components.
2:30 p.m.

Hardware Accelerators

Team Members: Ethan Clapp, Seth Gautreaux, Charles McDonald, Andrew Turner, William Wilson

Advisor: Dr. Miguel Gates

Field Programmable Gate Arrays (FPGAs) are integrated circuits functioning like “mini-computers”. One of the perks of FPGAs is their programmability, allowing for many different use cases. These devices can be programmed to perform tasks that are considered computationally expensive, a concept called hardware acceleration. Hardware acceleration is useful for systems that may not have a lot of hardware resources to dedicate to harder tasks. FPGAs are ideal, as they consume relatively little power for their performance. This project focuses on simulating this hardware acceleration; in particular, we set up a file server that utilizes an FPGA to accelerate AES encryption.
3:00 p.m.

PHOEBUS Heralds Operationally Enhanced Barriers for User Security (PHOEBUS)

Team Members:  Travis Knippers, Jacob Latino, Blake Perrin, Dakota Suire, Cameron Thomas

Advisor: Dr. Miguel Gates

PHOEBUS is a small office/home office (SOHO) router that performs local analysis on the entire network of connected devices. It employs traditional analyzation and packet capture technologies like WireShark and a machine learning (ML) model to provide network administrators with all of the information needed to discover potentially compromised systems. These solutions are nicely packaged behind a user-friendly GUI. PHOEBUS was made accessible to hobbyists, casual users, and small businesses and is deployable on older hardware.