M.S. Mechanical Engineering, Robotics
Carnegie Mellon University, 2025
Email: pdreyer@cmu.edu
Veteran - USAF
M.S. Mechanical Engineering, Robotics
Carnegie Mellon University, 2025
Email: pdreyer@cmu.edu
Veteran - USAF
I'm a Senior studying Mechanical Engineering and Robotics at Carnegie Mellon University. Along with my studies, I have completed an ML/Vision internship with Ford Motor Company, and participated in research with the Biorobotics Lab (Carnegie Mellon) and the Matter Assembly Lab (CU Boulder).
My biggest passions revolve around optimal control problems for mobile robot motion planning as well as computer vision development.
I'm a non-traditional student. Prior to academia, I was enlisted in the United States Air Force for six years, where I worked as a maintainer in radio frequency systems technology. I was fortunate enough to spend two years living and working in the United Kingdom, as well as 6-months deployed to Adana, Turkey. In 2020, I separated from the military as a staff sergeant (E-5) to pursue engineering and robotics full time.
PortfolioÂ
(Please see the menu in the top corner for pages with more details)
Here, I'm showcasing a demo of a hand-detection model being used to trigger interaction responses. "Comet" is a suspension-actuated mobile robot I'm developing with the goal of studying human-computer interaction at the embodied level. Comet is designed to be emotive and expressive in both how it shows attention, and it's responses to the user's gestures and voice commands. This method of actuation is both cost-effective (<500$) and feasible to operate in any room from 10'x10' to 100'x100'. I'm developing the full stack for this robot including mapping, motion planning / control, obstacle avoidance, sensor fusion / state estimation, UXUI, and hardware.Â
This is a toy example of a robot mapping an unknown environment using depth sensors (blue lines) that surround the robot in 3D. The robot uses the total entropy of the map to decide the optimal next action for exploring as fast as possible.
Here I showcase the use of reinforcement learning to navigate a planar workspace while avoiding singular robot configurations. the robot traverses a tight joint configuration which would ordinarily fall into a singularity.
I utilized a water jet to cut pieces of the frame from 16ga mile steel, which I then form with a hand brake to increase the rigidity and enable easier welding.
This XY-Gantry device enables users with tremors to create smooth handwriting or artwork. Through the use of a custom load cell, the entire user wrist interface acts as a force sensor to detect inputs from the user. With the combination of a low-pass filter applied to user force input signals, as well as limiting the acceleration seen at the motor output, we are able to significantly reduce the transmission of limb tremors to the writing interface. Provisional Patent.
(2023)
My team and I designed, analyzed, and manufactured a custom load cell for our tremor-filtering robotic device to allow for an elegant and cost-effective method of sensing user input.
(2023)
When building a lunar rover robot to compete in the NASA Artemis challenge, we needed a method of maneuvering through variable terrain conditions on the course. I designed and manufactured these wheels that allow us to remotely actuate the black grousers independently on each wheel. This allowed us to pivot from a highly efficient wheel for smooth terrain, to a high traction wheel for rough or granular terrain. (2021)
This is a design evolution of my prosthetic hand for below elbow amputees. The fingers on both devices are actuated by the user's elbow flexion, and provide haptic feedback to the residual limb via response from force sensors embedded in the finger tips. These devices show that low cost prosthetics can deliver high value functionality. IP donated to Humanos3D, a non-profit prosthetics company located in Medellin, Colombia. (2020)
This is a complete custom build of a stereo vision system for an ADAS system for my car. I utilize vehicle power from the 12V source within the car, as well as GigE ethernet to transmit data from the sensor payload to my host PC within my vehicle. This system allows me to fully utilize GPUs within the vehicle while lightweight hardware remains on the exterior of the vehicle. (2023)
ORB feature detection and FLANN matching are used here to extract velocity vectors of dynamic objects within a scene (vectors drawn for illustration). (2022)
Panel gaps are the thin spaces between exterior vehicle body parts. While uniform panel gaps typically indicate high-quality assembly of the car, nonuniform gaps usually indicate issues during the assembly of the vehicle that may also be found in critical operational components within. I developed a computer vision algorithm that allows a smartphone user to quickly inspect the panel gaps around a vehicle and observe exactly where issues exist, which may provide better new/used car purchasing insight. This video was used to discuss issues in thresholding that can cause problems with classical computer vision methods (2022)
Quadruped Fall Recovery - Independent Undergraduate Research, Carnegie Mellon University
Reachability Exploitation for Light Weight MPC - Independent Research
Remote Soft Actuator Position Sensing - MAC Lab, University of Colorado Boulder
Engineering Student of the Year - Pierce College (2020)
Provisional Patent - Tremor Filtering (2023)
Current (Fall 2025):
Modern Control Theory
Robot Dynamic Analysis
Robot Learning
How to Grow A Hardware Startup
Completed:
24370 Engineering Design 1
24321 Thermal Fluids Lab
15112 Introduction to Programming
24311 Numerical Methods
11344 Machine Learning in Practice
24678 Computer Vision for Engineers
24351 Dynamics
24352 Dynamic Systems + Controls
24322 Heat Transfer
76270 Writing for the Professions
21260 Differential Equations
21254 Linear Algebra and Vector Calculus
21111 Calculus I
21112 Calculus II
24262 Stress Analysis
24261 Statics
24231 Fluid Mechanics
24221 Thermodynamics
24200 Manual Machining
24100 Introduction to Mechanical Engineering
33141 Physics I
82131 Elementary Chinese I
09106 Modern Chemistry II
09105 Introduction to Modern Chemistry I
76260 Interpretation and Argument