Featured Projects

 

Machine Learning Course Projects

  • Exploratory Data Analysis for Machine Learning
    This project analyzes Paycheck Protection Program (PPP) loan data available through the Small Business Administration (SBA) through several visualizations and via testing of a hypothesis about loan forgiveness.

  • Supervised Machine Learning: Regression
    This project uses regression models on PPP loan data available through the SBA to predict the loan approval amount based on the available data and compare it against the real loan approval amount.

  • Supervised Machine Learning: Classification
    This project uses classification models on PPP loan data available through the SBA to predict if a loan was fully forgiven or not based on the available data and compares it against whether the loan was fully forgiven.

  • Unsupervised Machine Learning
    This project will be uploaded pending the public release of the data used.

  • Deep Learning and Reinforcement Learning

 
 
components.png

Soft Robotic Components

Soft robots require various sensing and actuating components that are made from soft or smart materials. It is important when designing soft systems, to understand how material properties can affect system performance as a whole. In my research, I’ve built sensors with liquid metal and exfoliated graphite composite and worked with various actuation methods, including shape memory alloys, McKibben actuators, and cables. I’m interested in deepening the community understanding of the materials used in soft robotics to help design systems.

 
fancy_view_spine.JPG

Modeling of a Highly-Coupled Cable-Driven Continuum Robot

Continuum robots have a wide range of applications from medical to space to general purpose robotic arms. These robots experience continuum deformations, like elephant trunks and snakes, rather than localized deformations at joints. Most cable-driven systems have used cable routing to decouple the robot’s segments, but there are advantages to having a system that is highly coupled, like biological systems.

 
skins.png

Robotic Skins

Robotic skins are 2D planar structures (generally elastomer films or fabrics) with sensing and actuating components. They can be wrapped around other passive bodies to impart motion to those bodies through surface manipulation. Robotic skins are easily reconfigurable and be used for a wide variety of tasks, such as manipulation and locomotion. These skins can also be used to assist in controlling the passive bodies.