About
Samuel Todd Flanagan
toddflanagan.web@gmail.com
I am an Engineering Scientist Associate at Applied Research Laboratories in Austin, TX. My current projects include GPU accelerated programing and deep learning.
Download my resume
Education
Texas A&M University
2019 - 2021
Master of Science in Computer Engineering
GPA 3.875
Texas A&M University
2013 - 2018
Bachelor of Science in Electrical Engineering
Minors in Computer Science and Mathematics
GPA 4.0
Publications
S. T. Flanagan. Signal Processing Improvements to Localization for Autonomous Vehicles. Master's thesis, Texas A&M University.
S. T. Flanagan, D. K. Khublani, J.-F. Chamberland, S. Agarwal, and A. Vora. Enhanced Normalized Mutual Information for Localization in Noisy Environments. 2020 IEEE Applied Signal Processing Conference (ASPCON), Kolkata, India, 2020, pp. 178-182.
S. T. Flanagan, D. K. Khublani, J.-F. Chamberland, S. Agarwal, and A. Vora. Localization in Autonomous Vehicles Using a Generalized Inner Product. 2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP), Ottawa, ON, Canada, 2019, pp. 1-5.
Experience
Applied Research Laboratories
Engineering Scientist Associate (Apr. 2021 – Present)
- Developed a deep learning ensemble framework with comparative metrics and model visualization
- Wrote a heap-based move median algorithm in CUDA C++, 400X speed up over original code
- Created a custom implementation of Score-CAM, an open-source model visualization package
- Advised team in machine learning capabilities, deficiencies, and underlying mathematics
Texas A&M University
Research Assistant (Spring 2019), Graduate Student Worker (Jan. 2019 – Jan. 2021)
- Proposed novel improvements to localization algorithms
- Developed a noise model for image acquisition in autonomous vehicles
- Wrote numerical simulations in Python to evaluate algorithmic improvements
Tanknology
Product Development Engineer (July – Nov. 2018)
- Wrote audio analysis software to detect leaks in VacuTect® recordings
- Integrated the software into the VacuTect® system to provide near-time auditory leak detection
- Wrote a UWP application for post-test leak detection in VacuTect® recordings (quality control)
- Built an Arduino based Sony VISCA camera controller for the TankCam® Inspection system
United States Department of Defense
Splunk Developer (May – Aug. 2017)
- Built multiple Splunk dashboards to visualize real-time data for organization managers
- Promoted data-driven decision making through the creation of effective Splunk dashboards
- Created multiple CSS files allowing developers to quickly improve dashboard design
musx
Product Development Intern (June – Aug. 2016)
- Learned cross-platform mobile development in JavaScript using React Native
- Developed a comprehensive mobile application beta test strategy
L-3 Communications
Electrical Engineering Co-op (May – Dec. 2015)
- Designed, fabricated, and installed instrumentation testing systems
- Solved unexpected problems with data acquisition equipment
- Contributed to the electrical design of aerospace mission systems
National Instruments
Test Engineering Intern (June – Aug. 2014)
- Updated production testing software for three adapter modules
- Tested updated software to verify functionality
Projects
IBM Senior Assistant
Electrical Engineering Senior Design Project (Fall 2017 – Spring 2018)
The IBM Senior Assistant is a prototype robot that helps elderly people live at home safely. It does this by following its user, performing fall detection, and reminding them to take their medications.
My contributions to the project:
- Android application
- Sends Wi-Fi credentials (inputted by the user) to the Raspberry Pi over Bluetooth
- Sends medication schedule information over Wi-Fi (TCP) to the Raspberry Pi
- Python code that receives Wi-Fi credentials from the app and connects the Raspberry Pi to the network
- TCP server (C++) that receives medication information from the app, generates a reminder message, and adds the reminder to Crontab (Linux scheduler)
- Node-RED flow that speaks out each reminder message at the proper time (IBM Watson Text-to-Speech)
- Python code that detects falls based on accelerometer readings
- Node-RED flow that asks the user if they need help after a fall is detected and sends a text message to get help if necessary (IBM Watson Assistant, Text-to-Speech, Speech-to-Text)
Link to code | Demo video
Hobbies
Rock Climbing
Outdoor sport climbing is my focus but great climbing is great no matter the style.
Abridged Tick List
- Poseidon, 5.13a (Belton Lake, TX)
- Super Cruiser, 5.13a (Reimers Ranch, TX)
- Atman, 5.10a Trad (Red Rock, NV)
- The Witch, V8- (Pedernales Falls, TX)
- DDD, V7 (Hueco Tanks, TX)
Running
Kolaches 4 a Cause
My younger brother and I planned this 50-mile charity run from Bastrop to Weikel’s Bakery in La Grange, TX.
We raised $4,524 for the Gazelle Foundation which will give about 150 people in Burundi, Africa access to clean water.
Disclaimer: I only made it 30 miles running and biked the last 20.