Academic Projects

Tsuneishi Holdings, Summer 2017

CERLAB Underwater Hull Cleaning Robot

Shipping vessels lose fuel economy when biofouling on their hull increases drag.  According to the Office of Naval Research, significant biofouling can account for up to 60% increases in drag, resulting in up to 40% increases in fuel costs*.  Current ship hull cleaning methods require dry-dock with manual spray at close range. 

​

To formulate an easier alternative, Tsuneishi Shipbuilding Co. teamed up with Carnegie Mellon University Professor Kenji Shimada and students in the Computational Engineering and Robotics Lab.  The robot is the subject of research for its novel under-actuated design.  

​

As a team member,  I am responsible for mobility controls and system identification.  The project requires strong working knowledge of Python, C++, and control theory.

​

Personal accompliments include:

• Testing of V3 Prototype at Tsuneishi Shipyard in Japan

• V3 Prototype control via XBox controller and IMU feedback in Python/C++ interface

• Interactive kyboard input mobility demo, including 6DOF motion and leg forces

• V3 Live IMU tracking animation

• Design for V4 Prototype

*Vietti, Peter (4 June 2009), New hull coatings for Navy ships cut fuel use, protect environment, Office of Naval Research, retrieved 21 May 2017

Mechatronics, Spring 2015

Automated Drill Tip Pre-Tinner

Sponsored by Kennametal, our team was tasked with processing tungsten-carbide drill tips and preparing them on a tinning tray.  Drill tips are currently sorted and processed by technicians which has led to repetitive strain injuries in the Kennametal workplace.  In industry, before the tips are attached to their respective boring bars or drills, silver wire pieces are partially melted on the upper tip surface.  These partially prepared tips can then be shipped anywhere in the world, ready to solder/insert into prefabricated drill slots. 

 Explicit system requirements:

• Ability to process different size/kerf drill tips

• Wiring connections must be permanent by final demo

• Tip placement on tray within array dimensions

• Fully autonomous, no manual intervention

Implicit system requirements​:

• Must process tips at reasonable rate (~9s per tip)

• Must easily allow flux refill as needed

• Should be modular for easy assembly/maintenance

Robotic Systems and the Internet of Things, Spring 2017

Selected IoT Projects

Under the direction of Professor Kenji Shimada, Carnegie Mellon University's Mechanical Engineering Department offered a new course in Spring 2017 allowing students to experiment with robotic systems and client-server frameworks.  Course topics included Simultaneous Localization and Mapping (SLAM), industrial robot communication, computer vision, and implementation of TSP algorithms.

Problem sets and course projects included:

• Python script control of DENSO 6DOF robot

• Pick and place operation with DENSO robot, webcam (OpenCV)

• Visual navigation of Raspberry Pi mobile car with PiCam (OpenCV)

• Path optimization for PiCar from realtime traffic data (use of AR markers in OpenCV)

• Two-story parking garage which actuates automatically, communicating with server database

Soft Robotics, Spring 2017

Soft Robotic Actuator

In the Spring of 2017, Soft Robotics students were given the opportunity to act on theory presented in class.  In most soft robotics systems, complex valve and accumulator banks are required for careful control of pressurization speed, limiting applications to lab environments. My project focused on the portability and reduced size of soft robotic actuators discussed during the semester.  

Using DIY silicon molding methods found online, our team was able to create a proof of concept system.  I focused on design and fabrication of the system while my partner, Matt Sahngjoon Lee was responsible for molding and testing Suzumori* style actuators and gripper end effectors.

​

*Original Suzimori Actuators

​

​