<Computer Science/>
Codes
In
ThinSat
DC Space Grant Consortium
Lily was under grant from the NASA DC Space Grant Consortium to develop a generic particle detector Arduino library for the ThinSat program. ThinSats are small satellites capable of transmitting data from low earth orbit. They are about the size of a piece of bread and are stacked into cubed ‘loaves’ (which are the size of a traditional CubeSat). ThinSat teams include middle school to college students and are managed by a lead institute such as a university or NASA site (Lily worked with American University’s ThinSat team). The goal of the ThinSat program is to promote STEM education by giving students a platform to develop satellite hardware, test sensor components with low and high altitude balloon flights, analyze data, and launch an actual payload into space.
ThinSats usually include a microcontroller (master) connected to a sensor component (minion). Most prebuilt sensor components come with a manufacturer-made software library ready for use with C++/Arduino, but if students want to build a sensor component or use a sensor component without a pre-made library, they must write their own. Writing a C++ software library can be daunting for middle and high school students or non-Computer-Science majors, so the goal of the particle detector library is to provide students with a generic minion library and corresponding base firmware (master) so they can have more freedom to build their own sensor components or use less documented sensor components.
The particle detector library is open-source and on Github. Lily worked on the base firmware code and the particle detector (minion) functions in Summer 2019 and the beginning of the Fall 2019 semester, then, the second half of the semester was spent configuring the I2C functionality between master and minion within the framework of the library.
Xinabox master (grey) connected to Teensy microcontroller minion (black) via AU-prototyped I2C board (copper)
A diagram of a ThinSat
NASA Partnership Provides Grants for AU Research
By Michael Valenti | July 8, 2019
From human-powered Mars rovers to space particle research, AU students and faculty are working on groundbreaking STEM projects thanks to funding from the DC Space Grant Consortium.
Read more.
DC Space Grant Consortium members at the 30th Anniversary Space Grant celebration at Capitol Hill (2020)
SpectraSearch
SpectraSearch is a full-stack web application that seeks to develop a database of spectral data and a searchable interface for spectral data/lights to help lighting professionals evaluate lighting products against current and future requirements. The interface will primarily allow users to search for lighting instruments by name or by spectral data, have access to spectral power distribution graphs, calculated values of metrics of interest such as CIE chromacity coordinates, CCT (correlated color temperature), response spectra convolution, and CRI (Color Rendering Index) General/Extended, and general information such as application, type, and manufacturer. The interface will also allow users to upload spectral and corresponding data to be approved by an administrator before being uploaded into the database; this feature includes a log-in function and admin page.
The open-source project was created by a small team of Software Engineering students at American University working with scientists at the Department of Energy’s Advanced Lighting Program as a primary stakeholder. The proof-of-concept software was debuted in a presentation to the DOE stakeholders for AU's Software Engineering final and includes a working front-end, with primary requirements such as search and scroll, connected to a SQL database, two color math backend functions (CCT and CRI), and Spectral Power Distribution graph display. The software primarily uses NodeJS, react.JS, D3.js, and SQL.
During the proof-of-concept run, Lily secured the stakeholders (see below) and remained the client liaison for the duration of the project, led and managed a team of six software engineering students including timeline design and sprint management, and served as the Scrum Master, the Database/Backend Engineer, and as a stand-in lighting SME.
During the Spring 2020 semester, the current team members continued to work with DOE stakeholders to put the proof-of-concept software into production as part of their Computer Science Senior Capstone course at American University. Plans for production include: populating the database, simplifying the spectral data column of the database, adding more color math functions, and a general production quality cleanup.
Lily was also invited to present a poster on the project at the 2020 DOE Lighting R&D Workshop co-sponsored by the Illuminating Engineering Society January 28–30, 2020.
2020 Team Members
Lily Donaldson | Client Liaison, Database/Backend Engineer, Scrum Master
Elisabetta Gabriele | Front-end Software Engineer
Rebecca Sakaguchi | UI/UX Engineer
Matthew Fields | Software Engineer (floating)
Faculty Advisor
Alex Godwin, Ph. D. | Software Engineering professor, Capstone Faculty Advisor
Stakeholders
Brian Walker, Ph. D. | DOE Lighting R&D Manager
Wyatt Miller, Ph. D. | AAAS Science & Technology Policy Fellow (DOE Lighting R&D Program)
Morgan Pattison | Senior Technical Advisor to DOE and principal of SSLS, Inc.
Arthur Shapiro, Ph. D. | American University Computer Science and Psychology Professor, Vision Scientist
Former Team Members
Anthony Baron, Chuyan Cheng, Yuhao Li, Zhiqi Lu
Screenshot of the SpectraSearch home page (click to enlarge)
Screenshot of search functionality
Screenshot of instrument page
ER Diagram
P5.js
P5.js is a JavaScript client-side library for creating graphic and interactive experiences. Lily was the Teaching Assistant to Professor Arthur Shapiro for the pilot of a lower-level Computer Science course using it (CSC 121 “Creative Coding”) at American University in Fall 2018. The course has since been integrated into the Computer Science department’s regular course offerings and approved to satisfy a General Education requirement for the university (Quantitative Literacy I). Lily helped test the course curriculum, wrote curriculum for a lesson on the World Wide Web and integrating P5.js with HTML/CSS into a Github Pages site, and wrote midterm and final revision sheets. She also used the Java version of P5.js, Processing, integrated with the SPRI lab’s Virtual Reality running room into a lesson on VR for Professor Shapiro’s Spring 2019 first year seminar “Reality: Distorted, Augmented, or Otherwise” where she was the course Program Leader.
Predictive Policing Module
An Example of Average Kerneling and Heat Mapping for use with Predictive Policing in P5.js
How it works:
- Crime data is taken in and cleaned for use. For this example, open source crime data from Cary, North Carolina was used. The data was stripped to crime and latitudinal and longitudinal location.
- The crime data is scaled to the drawing canvas and plotted by location.
- An averaging kernel algorithm is run. Each kernel is a ‘square’ which measures how many data points are within its boundaries. The kernels overlap (in this case an 80px by 80px kernel is placed every 10px on both the x and y axes to allow for detailed overlap). Each kernel’s display color is scaled based on the number of data points it contains. This display kernel has a low alpha value so that where the kernels overlap, an average crime rate (or color) is produced. This is a heat map.
- The user may create another averaging kernel by clicking on the screen. The user’s kernel checks each pixel’s color in the 30px by 30px box the user selects and averages the color (or crime rate) for the entire box. This color is displayed (sans the alpha value) in the box.
- The crime rate from within the user’s kernel is calculated and compared to the average crime rate for the entire map (the percent difference from the average). Negative values indicate a crime rate lower than the average crime rate, and positive values indicate a crime rate higher than the average crime rate. If the percent difference is lower than -105%, this means that the user’s kernel’s crime rate is insignificant or that the user’s kernel is out of bounds of Cary, and an error message is displayed.
Future work could include a method for separating based on crime type, ability to go by street/block level, and integration with a mapping API. This project was submitted for Dr. Bill Davies’ (Justice, Law, and Criminology) Spring 2019 Honors course, “The Legal Imagination,” which focused on law in works of fiction. The final project for the course was an essay or work of fiction on or inspired by one of the course materials. This project was submitted as an alternative to a written work (with Dr. Davies blessing) using “Minority Report” and the surrounding materials on predictive policing as its chosen course materials. P5.js was used because it is inherently graphics based and simple to understand in terms of programming languages; this was important because the project audience was a non-CS professor and a class of mostly non-STEM majors.
Reality: Distorted, Augmented, or Otherwise
First Year “Complex Problems” Seminar
Professor: Dr Arthur Shapiro; Co-Curricular Program Leader/Teaching Assistant: Lily Donaldson
Co-Curricular 1: The Pulfrich Effect – Illusions and Virtual Reality
Above: Students observe the Pulfrich Effect in 2D (with glasses)
Below Left: A student observes the Pulfrich Effect in Virtual Reality
Below Right: Simplified 2d Pulfrich Effect example (P5.js sketch)
Co-Curricular 2: Sheep Brain Dissection
Students dissected sheep brains in groups of three as part of the course’s Neuroscience / Psychology section.
Co-Curricular 3: Pokemon Go
Students played Pokemon Go (accomplishing a list of tasks in the game) using Augmented Reality as part of the course’s Augmented / Virtual Reality section. Students wrote short essays on an AR topic as an alternative to gameplay.
Websites
Website for construction company built in exchange for a $1000 donation to Children’s Miracle Network Hospitals (see Service for more information)
American University SPRI (Sensation, Perception, Reality, and Illusion) Lab Website. Fun fact: the dog mascot is Dr. Shapiro’s beloved pup, Wabbit, and along with building the site, Lily did all of the Wabbit illustrations!
DC Food Grants Honors Challenge Course Fall 2019. Faculty Advisor: Dr. Alex Godwin. The module was built in response to a need in the DMV (DC, Maryland, Virginia) community for a centralized food insecurity grant database. The Challenge Course project included academic research on food insecurity in the DMV and interviews with stakeholders in the database development. The database/module was built for non-coders to maintain, so Google Forms and Google Sheets were chosen for their simplicity and popularity. The module pings the Google preformatted Google Sheet (which can take data from a Google Form) via the Google API and returns a searchable and scrollable HTML table via NodeJS. The module will likely be handed off to the initial stakeholder, the DC Food Policy Council.
Miscellaneous
Final Project for CSC 330 OoCS (Organization of Computer Systems)
This is a game written entirely in Assembly language (because at the AU CS department, we like to party like it’s 1965). The game features a sprite inspired by Tamagotchi. The hardware includes a microcontroller, LED screen, potentiometer, and a piezo. The user can use the potentiometer to select “eat” or “love”. “Eat” feeds the sprite clovers, and “Love” makes the sprite’s eyes flash with hearts. When the user successfully selects “Eat” or “Love,” the piezo beeps.
Check out the code repository on Bitbucket (if you dare).
