DeepAI
Log In Sign Up

ContactNets: Learning of Discontinuous Contact Dynamics with Smooth, Implicit Representations

09/23/2020
by   Samuel Pfrommer, et al.
0

Common methods for learning robot dynamics assume motion is continuous, causing unrealistic model predictions for systems undergoing discontinuous impact and stiction behavior. In this work, we resolve this conflict with a smooth, implicit encoding of the structure inherent to contact-induced discontinuities. Our method, ContactNets, learns parameterizations of inter-body signed distance and contact-frame Jacobians, a representation that is compatible with many simulation, control, and planning environments for robotics. We furthermore circumvent the need to differentiate through stiff or non-smooth dynamics with a novel loss function inspired by the principles of complementarity and maximum dissipation. Our method can predict realistic impact, non-penetration, and stiction when trained on 60 seconds of real-world data.

READ FULL TEXT
11/15/2021

An Adaptive Framework for Reliable Trajectory Following in Changing-Contact Robot Manipulation Tasks

We describe a framework for changing-contact robot manipulation tasks th...
06/21/2021

Towards a Framework for Changing-Contact Robot Manipulation

Many robot manipulation tasks require the robot to make and break contac...
02/20/2021

Robotic Contact Juggling

We define "robotic contact juggling" to be the purposeful control of the...
07/02/2020

ADD: Analytically Differentiable Dynamics for Multi-Body Systems with Frictional Contact

We present a differentiable dynamics solver that is able to handle frict...
03/29/2021

Fundamental Challenges in Deep Learning for Stiff Contact Dynamics

Frictional contact has been extensively studied as the core underlying b...
09/11/2021

Bundled Gradients through Contact via Randomized Smoothing

The empirical success of derivative-free methods in reinforcement learni...
12/09/2019

A parallel-GPU code for asteroid aggregation problems with angular particles

The paper presents a numerical implementation of the gravitational N-bod...