Hello!
I'm Tom Gao. I am a current Master of Science - Robotic Systems Development (MRSD) student at Carnegie Mellon University graduating in May 2026. My academic interest is applying learning-based methods and transformer models to autonomous robotic tasks.
During undergrad, I majored in Computer Science and minored in Electrical Engineering, and was also highly involved with the Robotics Institute at University of Michigan. I assisted research in the Lab for Progress run by Prof. Chad Jenkins with a focus on using Neural Radiance Fields to learn and render multi-jointed articulated objects. I helped develop the initial ROB 102: Introduction to AI and Programming course (now a part of the Distributed Teaching Collaborative as HelloRob) with Prof. Chad Jenkins and Jana Pavlasek, the ROB 310: Signals and Sensors course with Dr. Peter Gaskell, and have worked since on software development for the MBot.
I had the opportunity to work with the Venomoth Team (Previously the PAX Team) at Amazon DTW10 in 2024, and delivered key solutions for vendor authorization infrastructure serving 3+ million vendors on Amazon.com. I learned a lot about delivering changes safely at scale, deep diving into the context of previous infrastructure designs and implementations, and designing and driving large projects to completion.
As part of the capstone project for the MRSD program, I am working on the Vision-based Autonomous DExtrous Reaper (VADER) project with four others, focusing on bimanual autonomous harvesting of green bell peppers in the field. As the implementation of the project advances in 2025, more details will be provided here.
Publications and Other Works
Dongjun Hwang, Sungwon Woo, Tom Gao, Raymond Luo, and Sunghwan Baek, "Invisible Watermarks: Attacks and Robustness". [ArXiv] (https://arxiv.org/abs/2412.12511)
Stanley Lewis, Tom Gao, and Odest Chadwicke Jenkins, "NARF24: Estimating Articulated Object Structure for Implicit Rendering". [ArXiv] (https://arxiv.org/abs/2409.09829)
Stanley Lewis, Tom Gao, Nick Janne, and Odest Chadwicke Jenkins, “Inverting the Design of Everyday Things: Affordances, Signifiers and Why Objects Should See and Robots Should Do.” NeuRL-RMW: Workshop for Neural Representation Learning for Robot Manipulation @CoRL2023, (https://neurl-rmw.github.io/) (accepted for oral spotlight)
Peter Gaskell, Jana Pavlasek, Tom Gao, Abhishek Narula, Stanley Lewis, Odest Chadwicke Jenkins. “MBot: A Modular Ecosystem for Scalable Robotics Education”, submitted to ICRA 2024. [ArXiv preprint] (https://arxiv.org/abs/2312.00962)