Graph network simulators can learn discontinuous, rigid contact dynamics

KR Allen, TL Guevara, Y Rubanova… - … on Robot Learning, 2023 - proceedings.mlr.press
Recent years have seen a rise in techniques for modeling discontinuous dynamics, such as
rigid contact or switching motion modes, using deep learning. A common claim is that deep …

Learning rigid dynamics with face interaction graph networks

KR Allen, Y Rubanova, T Lopez-Guevara… - arXiv preprint arXiv …, 2022 - arxiv.org
Simulating rigid collisions among arbitrary shapes is notoriously difficult due to complex
geometry and the strong non-linearity of the interactions. While graph neural network (GNN) …

Improvisation through physical understanding: Using novel objects as tools with visual foresight

A Xie, F Ebert, S Levine, C Finn - arXiv preprint arXiv:1904.05538, 2019 - arxiv.org
Machine learning techniques have enabled robots to learn narrow, yet complex tasks and
also perform broad, yet simple skills with a wide variety of objects. However, learning a …

Validating robotics simulators on real-world impacts

B Acosta, W Yang, M Posa - IEEE Robotics and Automation …, 2022 - ieeexplore.ieee.org
A realistic simulation environment is an essential tool in every roboticist's toolkit, with uses
ranging from planning and control to training policies with reinforcement learning. Despite …

Reactive planar manipulation with convex hybrid mpc

FR Hogan, ER Grau… - 2018 IEEE International …, 2018 - ieeexplore.ieee.org
This paper presents a reactive controller for planar manipulation tasks that leverages
machine learning to achieve real-time performance. The approach is based on a Model …

Parameter and contact force estimation of planar rigid-bodies undergoing frictional contact

N Fazeli, R Kolbert, R Tedrake… - … International Journal of …, 2017 - journals.sagepub.com
This paper addresses the identification of the inertial parameters and the contact forces
associated with objects making and breaking frictional contact with the environment. Our …

Fundamental limitations in performance and interpretability of common planar rigid-body contact models

N Fazeli, S Zapolsky, E Drumwright… - Robotics Research: The …, 2020 - Springer
The ability to reason about and predict the outcome of contacts is paramount to the
successful execution of many robot tasks. Analytical rigid-body contact models are used …

Learning data-efficient rigid-body contact models: Case study of planar impact

N Fazeli, S Zapolsky, E Drumwright… - … on Robot Learning, 2017 - proceedings.mlr.press
In this paper we demonstrate the limitations of common rigid-body contact models used in
the robotics community by comparing them to a collection of data-driven and data-reinforced …

Long-horizon prediction and uncertainty propagation with residual point contact learners

N Fazeli, A Ajay, A Rodriguez - 2020 IEEE International …, 2020 - ieeexplore.ieee.org
The ability to simulate and predict the outcome of contacts is paramount to the successful
execution of many robotic tasks. Simulators are powerful tools for the design of robots and …

Scaling Face Interaction Graph Networks to Real World Scenes

T Lopez-Guevara, Y Rubanova, WF Whitney… - arXiv preprint arXiv …, 2024 - arxiv.org
Accurately simulating real world object dynamics is essential for various applications such
as robotics, engineering, graphics, and design. To better capture complex real dynamics …