- Leading firmware design & development for low cost underwater gliders

- Founded a robotics team that competes on the Discovery Channel's TV show Battlebots- Leads engineering & manufacturing efforts to create “Bloodsport”, a 250lb robot - Responsible for chassis, drive and redundant electrical design of the robot. - Manages a team of 8 people to design, fabricate and compete once a year. - Consistently top performing, 70% winning record. Ranked #2 in the 2020 season. - Learn more here: www.bloodsportbattlebot.com

- Architected & prototyped software for a 3D printer fleet which utilized a collaborative robotic arm.- Led R&D software design and implementation for a new “automated 3D printer” product line.- Designed & integrated algorithms for improving mixer reliability to >97% while improving speed by 3x.- Implemented STM32 C firmware for limiting stepper motor vibrations by 5%, which enhanced print quality

- Designed framework for controlling multiple motors for use on high capacity train lifts in Python.- Implemented framework on several train lift systems currently in use.- Integrated collision avoidance and localization for carts carrying train wheels in ROS.- Designed and tested prototype for omnidirectional train transfer tables.

I helped develop the uPoint Mult-Robot Controller (MRC) for Defense and Security robots at iRobot. This controller is an android application where I implemented several features and fixed many bugs. In addition, I developed a smoke tests framework in Java that verified the application was functioning correctly. Part of these smoke tests was an experimental vision system that allowed the tests to confirm the robot was calibrated correctly and moved correctly, this was written in Python.

Developed a car detection and tracking algorithm for Velodyne LIDARs that were mounted on the bumpers of vehicles. This algorithm implemented binary tree searches for k-NN point cloud searches to not only group objects but to also determine contours along the vehicle. Once the vehicle could be identified as a Motorcycle, Car, Truck, or Semi using those methods, a Kalman filter was implemented to track the vehicle as it drove by the car, giving position and velocity estimations. The entirety of this algorithm was used to give ground truth to blind spot sensors and validate their functionality. Developed an environmental testing GUI which allowed for long-term validation of blind-spot sensors. This involved communicating with the sensors over CAN and using a NIDAQ system for detecting/measuring current. This system recorded information and automatically detected failures and recoveries.

I worked on developing a gripper system for an industrial robot named "Baxter" by Rethink robotics. The purpose of the system was to bag medicine cups and place them in a box. The solution to this to machine a gripper that could pick up a stack of about ten medicine cups and flip them upside down using a stepper motor and then drop them in a fixture that would then drop the cups into an industrial bagger. Once the bagger sealed the bag with the cups contained inside it, Baxter then went and picked up the bag and placed it into the box.Skills Learned: AutoCAD, Machining using CNC's and Lathe's, Electrical Design, Common Industrial practices