Michael Farrell
Staff Software Engineer, State Estimation
Carlsbad, California, United States
Summary
👤
Senior🎓
Top SchoolMichael Farrell is a Staff Software Engineer specializing in vision-based state estimation with nine years of experience building SLAM and odometry systems for aerial robots. Currently at Anduril after leading state estimation development for Shield AI’s autonomous quadrotor, he combines C++ systems engineering with research-grade algorithms to enable GPS-denied navigation and mapping. His graduate research produced a practical method for robust multirotor landings on moving platforms, showing measurable improvements over prior approaches and later informing fielded autonomy stacks. An active contributor to vision ML tooling—having worked on SuperPoint-related notebooks and descriptor evaluation—he bridges neural feature research and real-time robotics implementations. Based in Carlsbad, CA, he favors hands-on problem solving and field testing as much as code, and can often be found debugging systems between tacos on the beach.
10 years of coding experience
5 years of employment as a software developer
Master of Science - MS, Mechanical Engineering, Master of Science - MS, Mechanical Engineering at Brigham Young University
Spanish
Stackoverflow
31reputation
7kreached
2answers
0questions
Github Skills (12)
computer-vision10
jupyter-notebook10
opencv10
python10
image-processing9
data-visualization8
data-visualisation8
data-visualizations8
linux6
dynamic-linking6
boost6
gcc6
Programming languages (7)
Github contributions (5)
rpautrat/SuperPoint
Mar 2019 - Mar 2019
Efficient neural feature detector and descriptor
Role in this project:
Full-stack Developer
Full-stack DeveloperContributions:11 commits, 1 PR, 9 comments in 6 days
Contributions summary:Michael primarily contributed to the development of a notebook for viewing COCO exports, indicating a focus on data visualization and analysis within the context of the SuperPoint project. The user's changes involved modifications to existing notebooks and the inclusion of libraries for image manipulation and plotting, suggesting contributions related to data handling and feature extraction. They further modified code within a file related to descriptor evaluation. Additional commits appear to address model loading and matching descriptors.
deep-learningdescriptordetector
byu-magicc/roscopter
Jun 2017 - Dec 2019
*Under Development* - A fully-featured multirotor autopilot for ROS
Contributions:64 commits, 10 PRs, 43 pushes in 2 years 6 months
ros
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.