Internet Affordability

University of California, Santa Barbara

Objective:

Develop an interactive tool to visualize disparities in broadband plans offered by major United States Internet Service Providers

Method:

A Machine Learning algorithm was used to scan and collect pricing data across various addresses in 27 U.S. cities. The dataset was pre-processed by removing specific address information to ensure privacy, replacing it with geographic coordinates while maintaining location accuracy.

In each city, between one and three service providers, including Frontier, Spectrum, and AT&T, were analyzed to provide comparative pricing insights. The interactive heat map enables users to explore broadband pricing disparities, offering valuable insights into the relationship between pricing, city location, and service provider.

Explore the Heat Map

Children's Storybook Illustrator

University of California, Davis.

Objective:

"With the rapid advancement of technology, it is important to use its power in ways that benefit society. HackDavis challenges its participants to create for social good and create an opportunity for us to explore the intersection between technology and society." (https://hackdavis.io/) University of California, Davis.

Method:

For this project my team developed an interactive website that utilizes a machine learning algorithm to generate images for children's short stories. Our personal goal was to provide K-12 students with a tool that encouraged learning and sparked creativity, ultimately enhancing their educational experience.

Virtual Reality Walking Simulator

University of California, Santa Barbara.

Objective:

Develop a virtual walking simulator customized for individuals with limited mobility, offering immersive experiences designed to enhance accessibility and foster empowerment and inclusion.

Method:

While at UC Santa Barbara, I was involved in a research project at the Human-AI Integration Lab (HAL), which focused on developing a virtual walking simulator.

In this role, I honed my skills in Unity and Meta's Oculus to develop immersive virtual environments.

Feedback from our lead student researcher, who is paraplegic, was utilized to evaluate the comfort of using virtual tools for individuals new to walking experiences.

Drone Path Planning

University of California, Merced.

Objective:

Automate the process of piloting a drone through an obstacle course.

Method:

We were separated into teams of four with revolving roles. The "Controller" was responsible for plotting a course through the obstacle course using a controller, while another team member was tasked with retrieving the drone if it fell. The third position, "Manager," involved facilitating communication among team members. The final position was "Programmer," which I initially filled. We only had a brief time in each position, but my contributions laid the groundwork for our team's success.

To begin, I outlined potential functions we would require, such as "flying through hoops" and "landing." From there, my team seamlessly fleshed out the functionalities. It was rewarding to learn that our careful code organization was recognized as the key to our victory at the award ceremony.

California Wildfires

California State University, Stanislaus.

Objective:

Develop machine learning algorithms to identify fuel sources on the ground using aerial and satellite imagery to create better models of wildfires and track their atmospheric effects in California.

Method:

Using QGIS (Quantum Geographic Information System), a team of physics and computer science majors tagged fuel sources, including dry/wet grass, dead/live trees, houses, and non-burnable items like roads, rocks, and dirt. Since the team had already developed a MLP, my main role was to enhance the code's readability and accuracy.