I am a research scientist at NVIDIA, working on physics simulators and their applications in robotics. My research is focused on reducing the sim2real gap. I am investigating simulators as semantic models of the world that can be updated from measurements. Such models are parameterized by physically meaningful variables and can enable robots to perform complex tasks.
I obtained my Ph.D. degree in Computer Science from the University of Southern California (USC) advised by Gaurav Sukhatme and was supported by a Google PhD Fellowship. I received a Master’s degree from USC, and a Bachelor’s degree from the University of Rostock (Germany).
Interests
- Robot learning
- Physics simulators
- Autonomous robots
- Task and Motion Planning
Education
PhD Computer Science, 2022
University of Southern California
MSc Computer Science (Intelligent Robotics), 2017
University of Southern California
BSc Computer Science, 2015
University of Rostock, Germany
Projects
Selected Publications
Grasp'D: Differentiable Contact-rich Grasp Synthesis for Multi-fingered Hands
Synthesize contact-rich grasps for multi-fingered hands via differentiable simulation
Inferring Articulated Rigid Body Dynamics from RGBD Video
We present a pipeline to learn simulators from depth or RGB video. The “URDF” of a mechanism is reconstructed, and the simulation parameters are inferred through Bayesian inference.
Probabilistic Inference of Simulation Parameters via Parallel Differentiable Simulation
Reducing the reality gap through a Bayesian inference algorithm that leverages massive GPU parallelism and differentiable simulators.
Journal Articles
(2021).
NeBula: Quest for Robotic Autonomy in Challenging Environments; TEAM CoSTAR at the DARPA Subterranean Challenge.
Journal of Field Robotics.
(2020).
Bench-MR: A Motion Planning Benchmark for Wheeled Mobile Robots.
IEEE Robotics and Automation Letters (RA-L) 2021 and presented at ICRA 2021 and the ICAPS 2021 Workshop on Planning and Robotics (PlanRob).
(2019).
Scaling simulation-to-real transfer by learning composable robot skills.
International Journal of Robotics Research.
(2019).
Confidence-rich Grid Mapping.
International Journal of Robotics Research.
Conference Proceedings
(2019).
LAMP: Large-Scale Autonomous Mapping and Positioning for Exploration of Perceptually-Degraded Subterranean Environments.
International Conference on Robotics and Automation (ICRA) 2020.
(2019).
Physics-based Simulation of Continuous-Wave LIDAR for Localization, Calibration and Tracking.
International Conference on Robotics and Automation (ICRA) 2020.
(2019).
Auto-conditioned Recurrent Mixture Density Networks for Learning Generalizable Robotic Manipulation Skills.
Southern California Robotics Symposium (SCR).
(2018).
Heterogeneous Sensor Fusion via Confidence-rich 3D Grid Mapping: Application to Physical Robots.
International Symposium on Experimental Robotics (ISER).
(2018).
Scaling simulation-to-real transfer by learning composable robot skills.
International Symposium on Experimental Robotics (ISER).
(2018).
Gradient-Informed Path Smoothing for Wheeled Mobile Robots.
In International Conference on Robotics and Automation (ICRA).
(2017).
Confidence-rich Grid Mapping.
International Symposium on Robotics Research (ISRR).
(2017).
Planning High-speed Safe Trajectories in Confidence-rich Maps.
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).
(2017).
Concept and Realization of a Diagnostic System for Smart Environments.
International Conference on Agents and Artificial Intelligence (ICAART).
Workshop articles & Abstracts
(2020).
Sparse-Input Neural Network Augmentations for Differentiable Simulators.
DiffCVGP Workshop.
(2020).
Closing the Sim2Real Gap using Invertible Simulators.
RSS Pioneers 2020 Workshop.
(2020).
Augmenting Differentiable Simulators with Neural Networks to Close the Sim2Real Gap.
R:SS 2020 Workshop on Closing the Reality Gap in Sim2Real Transfer for Robotics.
(2019).
Real2Sim Transfer using Differentiable Physics.
R:SS Workshop on Closing the Reality Gap in Sim2real Transfer for Robotic Manipulation.
Experience
Research Intern
NASA Jet Propulsion Laboratory (JPL)
Researched representations and algorithms enabling autonomy of novel robot platforms.
Software Engineering Intern
Two Sigma Investments
Added support for parallel computing platform Apache Spark to the data science toolbox Beaker Notebook (now BeakerX).
Research Assistant
Information Sciences Institute (ISI)
Built large-scale information retrieval system for gaining information about >300k firms from company web pages.
Software Engineering Intern
Microsoft
Built business intelligence dashboard and language processing tools for software engineering teams.
Research Assistant
University of Rostock
Designed software tools and assisted with research projects in ubiquitous computing, bioinformatics, and web services.