As a member of the Personal Robotics Group at OSU, I am building a ROS-based framework for controlling and monitoring multiple robots over long periods of time (days to weeks) without direct human supervision. Intended features include a web-based GUI, the ability to perform multiple concurrent tasks on a single robot, and distributing tasks across a group of robots from a central controller.

As a software intern at PrecisionHawk, I worked on the development team for the PrecisionMapper website. Working on this software stack exposed me to Ruby on Rails, ReactJS, and techniques for interfacing with AWS S3 and EC2. Additionally, I worked to apply Docker, Vagrant, and Docker Compose to simplify configuring my local development environment, and built a small app for internal testing with Electron, React, and Redux.

Working in ARoS with Dr. Lobaton, I've had the opportunity to work on the following projects:Developed a framework for centrally commanding multiple Rolling Spider quadcopters using Bluetooth Low Energy. Initially, work focused on reliably calibrating and launching multiple drones simultaneously. The long term goal of this research is the use of the swarm as testbench for the sensors being developed for CINEMa (Cyborg Insect Networks for Exploration and Mapping)Additionally responsible for building a MATLAB simulation of the expected swarm dynamics, and transferring this behavior over to real-world testing.Simulated effectiveness of biologically inspired algorithms for localizing swarms of robots using radio signal strength (RSSI). After simulation, these methods were tested experimentally using a Mindstorms EV3 robot.Implemented a Python-based TCP server for real-time streaming of RSSI readings from a Mindstorms EV3 robot over local networks.As part of a multi-lab collaboration, built a MATLAB framework for performing feature extraction on inertial measurement data from a cockroach mounted IMU. Trained multiclass machine learning models to recognize and predict various motion modes of roach locomotion using these features.Investigated methods for integrating previously identified cockroach motion modes into an Extended Kalman Filter to design an accurate pose estimation system for the roach.

Built a webapp to display the content of csv files in a logical, hierarchical format. Used a combination of html, css, JavaScript, and jQuery to build the application.Designed a web application to automatically generate and display the content of thousands of performance log files hierarchically, reducing the time spent manually updating results. Log files were parsed using Python, and HTML files were generated using the Jinja templating engine. Combined with client-side JavaScript, these files provide a layered graphical interface that allows users to quickly find the information they need, and investigate logs down to the needed level of detail.

Worked as a research assistant at the Center for Robotics and Intelligent Machines. As part of my responsibilities in CRIM, I assisted Matthew Cramer with his thesis research on artificial sensorimotor networks. My work in this role included debugging code written in Matlab and LabVIEW, reformatting data into a form usable for data mining, using the open source software WEKA to mine the test results for correlations in order to develop more accurate tracking algorithms.In addition to my role with Matthew Cramer, I also assisted Alex Youngs in developing a webpage based control interface for a medical simulation mannequin. I wrote a python webapp that ran on an embedded webserver in the mannequin, and provided a GUI to control the vital signs of the mannequin with a wireless connection.