Kevin S. Tracy

I am a PhD student in The Robotics Institute at Carnegie Mellon University advised by Zac Manchester. I am also a student researcher at [Google] Intrinsic, working with Stefan Schaal on contact-rich manipulation.

I did my BS in Mechanical Engineering at Rice University, and my MS in Mechanical Engineering at Stanford University with a concentration in dynamics. At Stanford I served as a teaching assistant to Stephen Rock for Control Design Techniques, and Mac Schwager for State Estimation and Filtering for Robotic Perception.

I have interned at SpaceX, Astranis, Lockheed Martin, and Maxar, primarily working on guidance, navigation, and control for spacecraft.

Email  /  CV  /  Google Scholar  /  Twitter  /  Github

profile photo

I'm interested in differentiable convex and non-convex optimization, motion planning, contact physics, and estimation. I like finding optimization-based solutions to new and old problems.

b3do Efficient Online Learning of Contact Force Models for Connector Insertion
Kevin Tracy, Zac Manchester, Ajinkya Jain, Keegan Go, Stefan Schaal, Tom Erez, Yuval Tassa
in review
website / arXiv

A method for learning contact force models from data in a way that scales very well on a GPU with a novel Linear Model Learning (LML) algorithm.

b3do ReLU-QP: A GPU-Accelerated Quadratic Programming Solver for Model-Predictive Control
Arun L. Bishop, John Z. Zhang, Swaminathan Gurumurthy, Kevin Tracy, Zac Manchester
in review
arXiv / code

An ADMM based QP solver that runs on a GPU and can handle very large MPC problems commonly found in robotics.

b3do Convex Quasi-Dynamic Simulation of Rigid Point Clouds with Torsional Friction
Kevin Tracy, Taylor Howell, Zac Manchester
IROS Workshop on Leveraging Models for Contact-Rich Manipulation, 2023

A quasi-dynamic rigid-body simulator for point clouds with tangential and torsional friction for each of the points. The result is stable and fast simulation with a QP solver.

b3do Differentiable Collision Detection for a Set of Convex Primitives
Kevin Tracy, Taylor Howell, Zac Manchester
ICRA, 2023
arXiv / code / video

A framework for computing differentiable collision information betweens a set of convex primitives by solving for the minimum scaling of each shape that results in an intersection.

b3do Robust Entry Guidance with Atmospheric Adaptation
Kevin Tracy, Giusy Falcone, Zac Manchester
AIAA SciTech Forum and Exposition, 2023

An updated version of the CPEG algorithm that allows for free-final time trajectories combined with an atmospheric density estimation scheme.

b3do Data-Efficient Model Learning for Model Predictive Control with Jacobian-Regularized
Dynamic Mode Decomposition

Brian Jackson, Jeong Hun Lee, Kevin Tracy, Zac Manchester
CoRL, 2022
OpenReview / arXiv / code

An improved version of dynamic mode decomposition that includes jacobian regularization from an a priori. dynamics model for learned dynamics in optimal control.

b3do A Square-Root Kalman Filter Using Only QR Decompositions
Kevin Tracy

arXiv / code

The classic square-root Kalman Filter significantly simplified through the use of QR decompositions.

b3do DiffPills: Differentiable Collision Detection for Capsules and Padded Polygons
Kevin Tracy, Taylor Howell, Zac Manchester

arXiv / code

DiffPills computes collision information between capsules and padded polygons by forming and solving differentiable quadratic programs.

b3do CALIPSO: A Differentiable Solver for Trajectory Optimization with Conic and Complementarity Constraints
Taylor Howell, Kevin Tracy, Simon Le Cleac'h, Zac Manchester
ISRR, 2022
arXiv / code

CALIPSO is a nonlinear programming solver designed to handle the types of conic and complimentarity problems that arise in robotic motion planning problems with contact.

b3do The V-R3x Mission: Towards Autonomous Networking and Navigation for Cubesat Swarms
Max Holliday, Kevin Tracy, Zac Manchester, Anh Nguyen
4S Symposium Small Satellites Systems, 2022

V-R3x is a three cubesat constellation that demonstrated on-orbit mesh networking and relative navigation.

b3do Ultra-Fine Pointing for Nanosatellite Telescopes With Actuated Booms
Kevin Tracy, Zac Manchester Ewan Douglas,
IEEE Aerospace Conf., 2022,   Best Paper (Avionics and Electronics for Space Applications)

A control and estimation framework is proposed for a fine pointing cubesat where traditional reaction wheels are swapped for three actuated booms with tip-mounted masses.

b3do CPEG: A Convex Predictor-corrector Entry Guidance Algorithm
Kevin Tracy, Zac Manchester
IEEE Aerospace Conf., 2022
pdf / code

CPEG is an online-capable tool for optimal entry guidance in the Martian atmosphere.

b3do Practical Limits on Nanosatellite Telescope Pointing: The Impact of Disturbances and Photon Noise
Ewan Douglas, Kevin Tracy, Zac Manchester
Frontiers in Astronomy and Space Sciences, Volume 8, 2021
pdf / arXiv

The environmental disturbances on a nanosatellite are analyzed within a novel control framework to establish practical performance limits.

b3do Low-Thrust Trajectory Optimization Using The Kustaanheimo-Stiefel Transformation
Kevin Tracy, Zac Manchester
AAS/AIAA Space Flight Mechanics Meeting, 2021
pdf / code

A trajectory optimization solver based on differential dynamic programming with an augmented lagrangian to handle conic constraints.

b3do Planning With Attitude
Brian Jackson, Kevin Tracy, Zac Manchester
IEEE RA-L, 2021,   Nominated for Best Student Paper

A unified approach to quaternion differential calculus for use in optimization.

b3do ALTRO-C: A Fast Solver for Conic Model-Predictive Control
Brian Jackson, Tarun Punnose, Daniel Neamati, Kevin Tracy, Rianna Jitosho, Zac Manchester
ICRA, 2021
pdf / code

A trajectory optimization solver based on differential dynamic programming with an augmented lagrangian to handle conic constraints.

b3do Model-Predictive Attitude Control for Flexible Spacecraft During Thruster Firings
Kevin Tracy, Zac Manchester
AAS/AIAA Astrodynamics Specialist Conference, 2020
pdf / code

Convex model-predictive control is used to stabilize a flexible spacecraft during planned thruster firings.