Carson Kohlbrenner

I am a PhD student with the Human Interaction & Robotics Group at the University of Colorado Boulder advised by Alessandro Roncone. My research interests lie in whole-body perception and learning for collaborative robotics. Specifically, my research is focused on establishing a body schema in robots through tactile sensing and artificial peri-personal space.

I am currently the primary maintainer of the GenTact Toolbox, a computational design pipeline for generating whole-body tactile skins that are fully 3D printable.

When I'm not immersed in research, you can find me enjoying a round of golf, producing music, or enjoying the outdoors. I am always interested in new opportunities and collaborations!

Email / CV / Google Scholar / Github / Medium / LinkedIn

Impacts

2026

Outstanding Paper Award by the University of Colorado Boulder for my latest ICRA paper — Outstanding Paper AwardCU Boulder Graduate Awards, Boulder, Colorado

2025

Live Demo at Dr. Roncone's Tenure Talk — Live DemoDr. Roncone's Tenure Talk, University of Colorado Boulder

2025

Community Recognition: Brett Adcock, X — Community Recognition (X)Brett Adcock, Founder and CEO of Figure AI

2025

Publications

	Design, Mapping, and Contact Anticipation with 3D-printed Whole-Body Tactile and Proximity Sensors IEEE International Conference on Robotics and Automation (ICRA), 2026 Carson Kohlbrenner, Anna Soukhovei, Caleb Escobedo, Nataliya Nechyporenko, Alessandro Roncone project page / code / arXiv
	GenTact Toolbox: A Computational Design Pipeline to Procedurally Generate Context-Driven 3D Printed Whole-Body Tactile Skins IEEE International Conference on Robotics and Automation (ICRA), 2025 Carson Kohlbrenner, Caleb Escobedo, S. Sandra Bae, Alexander Dickhans, Alessandro Roncone project page / code / arXiv
	A Machine Learning Approach to Contact Localization in Variable Density Three-Dimensional Tactile Artificial Skin NeurIPS - Workshop on Touch Processing, 2024 Carson Kohlbrenner, Mitchell Murray, Yutong Zhang, Caleb Escobedo, Thomas Dunnington, Nolan Stevenson, Nikolaus Correll, Alessandro Roncone code / arXiv
	A Sensor Position Localization Method for Flexible, Non-uniform Capacitive Tactile Sensor Arrays arXiv Preprint, 2024 Carson Kohlbrenner, Caleb Escobedo, Nataliya Nechyporenko, Alessandro Roncone code / arXiv

Technical Reports

	Whole-Body Dynamic Obstacle Avoidance with Humanoid Skin Niraj Pudasaini, Carson Kohlbrenner, William Xie Deep Reinforcement Learning Final Project, 2025 paper A recreation of the Bullet Time scene using a humanoid proximity sensing skin and a deep reinforcement learning agent.
	Converting the EuRoC MAV Dataset to MCAP and visualizing SLAM using Foxglove Carson Kohlbrenner Foxglove Tutorial, 2025 article / code A step-by-step guide to visualizing SLAM data using Foxglove.
	Training CLIP to Recognize Objects with Tactile Sensors Carson Kohlbrenner, Nikolaus Correll Transformer Based Robotic AI, 2024 article / code A step-by-step guide to training vision transformers to recognize objects with tactile sensors using contrastive learning.
	Self-Landing Rocket using Deep Reinforcement Learning Carson Kohlbrenner, Owen Craig, Thomas Dunnington Decision Making Under Uncertainty Final Project, 2024 paper / code Deep Q-network, behavior cloning, and heuristic PID control methods for safely landing a rocket on a platform are implemented and compared.
	Autonomous Localization for GPS-denied Aerial Tracking and Reconnaissance (ALLIGATR) Carson Kohlbrenner, Thomas Dunnington, Nolan Stevenson, Nyah Baltazar, Andrew Kabos, Marcus Quintanilla, Yarden Kelmann, Blair Schulze, Nicholas Grant, Zane Vandivere, Lucia Witikko, Brendan Bradley Senior Capstone Project, 2024 paper / code An autonomous drone equipped with two cameras for tracking moving targets with computer vision. Targets were successfully localized within 2 meters RMS.
	Linear Temporal Logic Planning for Quadcopter Reconnaissance in Densely Populated Environments Carson Kohlbrenner, Nolan Stevenson Motion Planning Final Project, 2023 paper / code Satellite images of densely populated forests are used to plan waypoints for a quadcopter to fly through while avoiding obstacles. Waypoints allow RRT to visit targets in a set order and avoid obstacles.

Site template built by Jon Barron.