B.S. Robotics Engineering
Michigan Technological University
(Dec. 2023)
Hunter Gipp
I am in my final semester of a Robotics Engineering Bachelor's degree (expected December 2023)
from Michigan Technological University. My primary interests are in automotive, robotics, and software.
My experiences have involved combining software and hardware components, defining appropriate
simulation requirements and coordinating the design, development, and deployment of integrated systems.
I have demonstrated leadership within projects and work,
with additional experience in project management.
This website provides an overview of the various work experiences, projects, and other related activities
that I have been a part of since the start of my bachelor's degree.
MathWorks Simulation Challenge, Simulation & Requirements Verification Team
About the challenge:
The AutoDrive Challenge II is a four-year competition hosted by GM and SAE in which college teams create and test a Level 4 autonomous vehicle.
Within AutoDrive lies the MathWorks Simulation Challenge, where teams utilize MathWorks tools such as
Matlab, Simulink, & RoadRunner to create and execute simulated scenarios to test components of the autonomous system.
Skills Applied:
What? Project Manager for the Simulation & Requirements Verification Team.
How? Used "OpenProject" management software to plan project timelines, allocate resources,
and track team progress.
Results: Gained leadership experience, as well as improving communication skills.
What? FOV analysis conducted in Matlab's Driving Scenario Designer.
How? Create simulation scenarios to test the range limits and field of view and boundaries of the car's sensors.
Results: Communicate results with perception & build teams to implement ideal sensor suite onto the vehicle.
What? Lateral & longitudinal testing of pure pursuit & PID controllers.
How? Create simulation scenarios that isolate lateral and longitudinal movements.
Results: Qualitative & quantitative results from simulation aide in tuning controller parameters.
What? Dynamic route planning and collision avoidance to evaluate planning algorithms.
How? A* search algorithm integrates directly into RoadRunner to find the optimal path from an HD map of the course.
Results: A* correctly reroutes when faced with random destinations and unexpected blockages.
2nd place: Concept Design Report & SRS
2nd place: Concept Design Event
3rd place: Overall Dynamic Challenges
5th place: MathWorks Simulation Challenge
U.S. Army Corps of Engineers, Construction Engineering Research Laboratory (CERL)
Skills Applied:
The Dynamic Cone Penetration Test is performed by driving a metal cone into the ground, repeatedly striking it with an 8 kg weight dropped from a distance of 575 mm. The Automated Dynamic Cone Penetrometer (ADCP) is an automated solution to this manual test. The ADCP utilizes a Python-based state machine implementation with ROS integration to ADCP, depth camera, and GPS.
A depth camera is attached to the ADCP to constantly track the ruts created by the front vehicle. When the rut passes a depth threshold, a message will be sent to the user to start a test. An interactive GUI displays rut depth and rut depth change, camera status, and current state. As the ADCP rod is hammered into the ground, the number of blows & depth in mm is displayed.
When a test is queued, the GPS location is pinged. Then, the ADCP hammers the rod into the ground, testing the Cone Index (CI) & Vehicle CI (VCI) coefficients of the soil to determine whether or not it's safe to continue onward. The location and results are saved to .CSV file and published. Additionally, a location based a-priori map is created with GO/NOGO indications.
Greenheck Group
Skills Applied:
High volume, low mix = automation opportunity!
Using two ABB robots and an automation friendly riveter,
the scope of this project is 442,118 WD-style blades per year,
covering 93.1% of all WD blades. This project reinforces the value
of continuous improvement, and increases safety in the workplace.
WD blades are picked from the Rollformer and placed on riveter pins along with hinges, to be joined. Upon completion of training in ABB RobotStudio, programming a proof of concept for this project began. The simulation and testing were used to verify positioning, accuracy, and speed of this automation process.
The final product produces 5 completed WD blades per minute, yielding a 30% increase in efficiency, as well as streamlining production. The process is also scalable to meet demand, and has the ability to surpass 2.5 million blades per year.
Robotic Systems Enterprise (RSE)
Skills Applied:
Using a Pioneer 3-AT robot, the goal was to create a robotic t-shirt cannon equipped with remote control driving, aiming, and firing; all via PS4 controller.
Brainstormed designs by conducting a design concept, using pugh analysis methods to compare and select attributes based on requirements and restrictions.
Converted the cannon fuel from CO2 gas to HPA using a 48ci/3000 psi paintball tank. This method allowed for 7 total shots, with distances up to 45 feet.