Senior Design Projects

ECS193 A/B Winter & Spring 2021

Database Design and App Development for Air Quality Research Center’s Fourier Transform Infrared (FT-IR) Spectroscopy Laboratory

Email **********
Ann Dillner
UC Davis

Project's details

Database Design and App Development for Air Quality Research Center’s Fourier Transform Infrared (FT-IR) Spectroscopy Laboratory
Dr. Dillner’s FT-IR laboratory currently employs three FT-IR instruments, operating semi-continuously, to analyze over 35,000 atmospheric fine aerosol (pollution) samples per year for clients that include the U.S. Environmental Protection Agency (EPA), U.S. National Park Service (NPS), the SPARTAN international monitoring network and the NASA MAIA satellite project. The data made available by these instruments supports research related to human health, visibility, and climate change. Each measurement from the instruments is inherently high-dimensional with over 10,000 variables per sample. While the development of an on-site database management system (DBMS) for routine analysis is on-going, we currently lack the expertise to migrate our file system to a DBMS for special research projects, capable of interacting with our research-level analysis tools developed over the past decade. Specifically, integrating our exploratory, multivariate regressions and QA/QC visualization into a user-friendly app to be used by staff will augment existing project design workflows and streamline aerosol research in Dr. Dillner’s laboratory.
Aim 1: A capable design team should revamp current database architecture (currently a file system) to handle raw instrument outputs, perform fundamental signal processing tasks, and store instrument outputs on secure local servers. Databases should be designed such that data acquisition, processing, and storage are robust to I/O errors that may occur during acquisition and analysis. Furthermore, database design and querying should be straightforward enough to have trained staff take corrective action if samples are for example, mislabeled or require removal from the database.
Aim 2: The design team works in collaboration with a postdoctoral researcher to migrate stand-alone multivariate analysis and machine learning tools onto a user-friendly app. The app should be able to access the database, perform the necessary analyses, visualize high-dimensional data, generate reports, and save analyses to secure servers. Ideally, the core functionality of the app (defaults) should be straightforward enough to use with minimal training.
1. Develop database management system (DBMS) for special research projects (50%).
a. Author a brief “How To…” manual to work in DBMS to be used to train and provide guidance to staff users of the DBMS.
2. Integrate multivariate analysis, machine learning, and QA/QC tools into DBMS accessible app (50%).
a. Analysis language must be written in R Programming language
b. Data visualization and a stable “default” configuration of multivariate regression a must
c. Ability to generate reports for laboratory clients
1. Experience in database development (in SQL) preferred
2. Experience with R programming language and machine learning
3. Experience with app development
**********
30-60 min every two weeks
Open source project
Attachment Click here
No
Team members N/A
N/A
N/A