Imperative learning: A self-supervised neural-symbolic learning framework for robot autonomy

C Wang, K Ji, J Geng, Z Ren, T Fu, F Yang… - arXiv preprint arXiv …, 2024 - arxiv.org
Data-driven methods such as reinforcement and imitation learning have achieved
remarkable success in robot autonomy. However, their data-centric nature still hinders them …

Port-Hamiltonian Neural ODE Networks on Lie Groups For Robot Dynamics Learning and Control

T Duong, A Altawaitan, J Stanley… - arXiv preprint arXiv …, 2024 - arxiv.org
Accurate models of robot dynamics are critical for safe and stable control and generalization
to novel operational conditions. Hand-designed models, however, may be insufficiently …

Approximation of nearly-periodic symplectic maps via structure-preserving neural networks

V Duruisseaux, JW Burby, Q Tang - Scientific reports, 2023 - nature.com
A continuous-time dynamical system with parameter ε is nearly-periodic if all its trajectories
are periodic with nowhere-vanishing angular frequency as ε approaches 0. Nearly-periodic …

Hamiltonian dynamics learning from point cloud observations for nonholonomic mobile robot control

A Altawaitan, J Stanley, S Ghosal… - … on Robotics and …, 2024 - ieeexplore.ieee.org
Reliable autonomous navigation requires adapting the control policy of a mobile robot in
response to dynamics changes in different operational conditions. Hand-designed dynamics …

Geometry preserving Ito-Taylor formulation for stochastic hamiltonian dynamics on manifolds

S Panda, A Gogoi, B Hazra, V Pakrashi - Applied Mathematical Modelling, 2023 - Elsevier
Naturally occurring systems often have inherent uncertainties and often evolve on
complicated smooth and differentiable hypersurfaces that are not necessarily Euclidean …

Practical perspectives on symplectic accelerated optimization

V Duruisseaux, M Leok - Optimization Methods and Software, 2023 - Taylor & Francis
Geometric numerical integration has recently been exploited to design symplectic
accelerated optimization algorithms by simulating the Bregman Lagrangian and Hamiltonian …

Towards enforcing hard physics constraints in operator learning frameworks

V Duruisseaux, M Liu-Schiaffini, J Berner… - ICML 2024 AI for …, 2024 - openreview.net
Enforcing physics constraints in surrogate models for PDE evolution operators can improve
the physics plausibility of their predictions and their convergence and generalization …

Projected Neural Differential Equations for Learning Constrained Dynamics

A White, A Büttner, M Gelbrecht, V Duruisseaux… - arXiv preprint arXiv …, 2024 - arxiv.org
Neural differential equations offer a powerful approach for learning dynamics from data.
However, they do not impose known constraints that should be obeyed by the learned …

PhysORD: A Neuro-Symbolic Approach for Physics-infused Motion Prediction in Off-road Driving

Z Zhao, B Li, Y Du, T Fu, C Wang - arXiv preprint arXiv:2404.01596, 2024 - arxiv.org
Motion prediction is critical for autonomous off-road driving, however, it presents significantly
more challenges than on-road driving because of the complex interaction between the …

The TILOS AI Institute: Integrating optimization and AI for chip design, networks, and robotics

AB Kahng, A Mazumdar, J Reeves, Y Wang - AI Magazine, 2024 - Wiley Online Library
Optimization is a universal quest, reflecting the basic human need to do better. Improved
optimizations of energy‐efficiency, safety, robustness, and other criteria in engineered …