Feedback Motion Planning for Liquid Transfer using Supervised Learning

TL;DR

This paper introduces a real-time motion planning algorithm for liquid transfer that leverages neural network-based system identification to handle high-dimensional fluid dynamics, achieving high success rates in simulation.

Contribution

The paper presents a novel fluid-aware motion planning method combining receding-horizon optimization with neural network-based system identification for high-dimensional liquid transfer tasks.

Findings

High success rates in simulated 2D and 3D fluid transfer benchmarks

Effective handling of high-dimensional configuration space

Real-time planning capability

Abstract

We present a novel motion planning algorithm for transferring a liquid body from a source to a target container. Our approach uses a receding-horizon optimization strategy that takes into account fluid constraints and avoids collisions. In order to efficiently handle the high-dimensional configuration space of a liquid body, we use system identification to learn its dynamics characteristics using a neural network. We generate the training dataset using stochastic optimization in a transfer-problem-specific search space. The runtime feedback motion planner is used for real-time planning and we observe high success rate in our simulated 2D and 3D fluid transfer benchmarks.