Education
Georgia Tech, Sept. 2020 - May 2021
Non-degree program
Coursework
Introduction to Operating Systems; Simulation, Probability, and Statistics; Deterministic
Stanford Center for Professional Development Sept 2019 - March 2020
Non-degree program
Coursework
- Principles of Robot Autonomy I. Topics: trajectory planning, Kalman filter, EKF, SLAM, ROS, autonomy stack
- Principles of Robot Autonomy II. Topics: computer vision, CNN, reinforcement learning, object grasping
Stanford University, Sep. 2011 - Jun. 2016
BS & MS in Mechanical Engineering, GPA 3.8/4.0
Concentration in Mechatronics
Relevant Coursework
Mechatronics; Vehicle Dynamics and Control; Artificial Intelligence; Feedback Control Design; Dynamics; Advanced Thermal Systems; Fluid Mechanics
Work Experience
Mainspring Energy - Menlo Park, CA | Feb. 2021 - Present
Controls Engineer
Startup (Stealth Mode) - San Francisco Bay Area, CA | Mar. 2019 - Feb. 2021
Mechanical Engineer
Led the design and integration of several key automation systems within a biotech instrument, which significantly improved reliability and serviceability. Wrote a number of Python scripts for system diagnostic checks, as well as manage our bill of materials.
SpaceX, Hawthorne, CA | Aug. 2016 - Aug. 2018
Structures Engineer
Worked as a mechanical design engineer on SpaceX’s Falcon 9 payload fairing. Was responsible for various secondary structure flight parts such as seals & gaskets, sensor mounts, TPS (thermal protection systems), and composite structures. Led the development of parts from concept and flight, which included CAD modeling, GD&T, FEM stress analysis, material strength testing, overseeing manufacturing/assembly, part acceptance/qualification testing, and resolving build issues. Lead meetings to review designs and identify issues or potential flight risks in order to develop solutions.
QBotix, Menlo Mark, CA | Jun. - Sep. 2014
Mechanical Engineering Intern
Developed methods and devices to test functionality and wear of a flexible shaft that transfers torque between a solar tracker and robot. This involved brainstorming failure modes and solutions, designing, manufacturing, and coding automated testing rigs.
Stanford Mechanical Engineering Department, Stanford, CA | Sep. 2014 - Jun. 2016
Teaching Assistant
Course assistant for an introductory fluid mechanics class and introductory thermodynamics class. Duties involve holding office hours, running experimental lab sections, and leading tutorial sections.
Projects
Multivariable Controls | May 2022
Title: Cycle-to-Cycle Control of an Opposed-Cylinder Free Piston Engine Using H-inf and µ-Synthesis Techniques
Demonstration of applying multivariable control techniques to control of an opposed-cylinder free piston engine. A continuous-time nonlinear plant model is developed, which is converted to discrete time to represent cycle-to-cycle dynamics. The discrete time plant is linearized to aid in controller development. Performance of controllers developed using H-inf, and µ-Synthesis, are compared to a linear-quadratic integral controller. Project report for coursework.
Visual Odometry | Aug. 2020 - Sept. 2020
Building algorithm for visual odometry based on material taught by University of Zurich. Topics used include camera calibration, SIFT feature detector and descriptor, RANSAC, stereo vision, Eight-point Algorithm, Lucas-Kanade optical flow tracker.
Turtlebot in ROS
Group project for class to implement full autonomy stack on Turtlebot to autonomously traverse across a map without colliding into obstacles. The project was completed in a simulated environment using ROS Gazebo. SLAM was implemented using ROS’s gmapping module with simulated LIDAR data, and path planning was implemented ourselves using A* and path smoothening. In addition, I implemented a exploration feature to allow the robot explore the unmapped territory until the complete closed area is mapped.
Self-Balancing Robot | Nov. 2018 - Mar. 2019
Self-balancing robot that is able to maintain its balance on two wheels and move forward, backwards, and turn. Goal was to program and design a feedforward/feedback control algorithm that is able to make the robot run more “sporty” than how these DIY robots are typically run. Fwd/aft movement in video is was commanded to show agility achieved. Commands to robot are controlled manually using joystick on a wireless RC controller. Prior to implementation, a model was run in Python to verify effectiveness of control algorithm. Hardware was purchased as off-the-shelf kit.
Details>>
Autonomous Game-Playing Robot | Feb. - Mar. 2016
Final team project for mechatronics class to create an autonomous robot that plays against another robot in a game to win the U.S. presidential election. The robots must navigate and traverse the map and "capture" more polling stations than its opponent before time is up to win the game.
Details >>
Game-Playing AI
A partner and I have coded a AI algorithm that plays the video game Dr. Mario, which is a game similar to Tetris. We have used Expecitmax to simulate the outcomes of future pills and evaluate the best action to take. For that evaluation process, we used a learning algorithm called TD-learning for the AI to teach itself what state of the board is more or less favorable. The AI performed better than a experienced human player.
Details >>
Autonomous Driving | May 2015
Team project for a vehicle dynamics class to write and test a control algorithm to drive a research vehicle along a predetermined path in a parking lot. Throughout its 250m drive, we were able to keep the vehicle within 0.2m of the desired path.
Details >>
Quadcopter | Jan. - Nov. 2013
A personal project I pursued to learn more about robotics, controls, and flight. I coded an Arduino from the ground-up to use as a flight controller for a quadcopter. This involved coding an attitude heading reference system using rotation matrices and accelerometer and gyro data, coding a controller for yaw, pitch, and roll stabilization.
Details >>
Skills
Software Engineering: C/C++, MATLAB, Python, Linux, Assembly Language (PIC Microcontroller)
Mechanical Engineering: GD&T, CAD (Solidworks/NX), Finite Element Stress/Modal Analysis (Nastran/Femap, Ansys)
Activities
Bike & Build | Jun. - Aug 2015
Biked across the U.S. from Maine to Santa Barbara with 31 other young adults to support affordable housing with Bike & Build, a non-profit organization. Out team raised a total of $168,000 and spent 15 days building with affordable housing organizations such as Habitat for Humanity. Strengthened my perseverance and teamwork skills by overcoming unexpected challenges.
Bio
Born in New York, raised in Tokyo, high school in Boston, and been in sunny California since then.
Interests: trail running, pickup soccer, mountain biking, photography, eating ramen
Favorite Movies: October Sky, Good Will Hunting, Back to the Future
yuto.watanabe@gmail.com | 978-806-5218 | Linkedin