Development of the Senseiver for efficient field reconstruction from sparse observations

Javier E. Santos, Zachary R. Fox, Arvind Mohan, Daniel O’Malley, Hari Viswanathan, Nicholas Lubbers

Research output: Contribution to journalArticlepeer-review

12 Scopus citations

Abstract

The reconstruction of complex time-evolving fields from sensor observations is a grand challenge. Frequently, sensors have extremely sparse coverage and low-resource computing capacity for measuring highly nonlinear phenomena. While numerical simulations can model some of these phenomena using partial differential equations, the reconstruction problem is ill-posed. Data-driven-strategies provide crucial disambiguation, but these suffer in cases with small amounts of data, and struggle to handle large domains. Here we present the Senseiver, an attention-based framework that excels in reconstructing complex spatial fields from few observations with low overhead. The Senseiver reconstructs n-dimensional fields by encoding arbitrarily sized sparse sets of inputs into a latent space using cross-attention, producing uniform-sized outputs regardless of the number of observations. This allows efficient inference by decoding only a sparse set of output observations, while a dense set of observations is needed to train. This framework enables training of data with complex boundary conditions and extremely large fine-scale simulations. We build on the Perceiver IO by enabling training models with fewer parameters, which facilitates field deployment, and a training framework that allows a flexible number of sensors as input, which is critical for real-world applications. We show that the Senseiver advances the state-of-the-art of field reconstruction in many applications.

Original languageEnglish
Pages (from-to)1317-1325
Number of pages9
JournalNature Machine Intelligence
Volume5
Issue number11
DOIs
StatePublished - Nov 2023

Funding

J.E.S. and Z.R.F. gratefully acknowledge the support of the US Department of Energy through the LANL/LDRD Program and the Center for Non-Linear Studies (CNLS) for this work. H.V. gratefully acknowledges primary support from the Department of Energy, Office of Science, Office of Basic Energy Sciences, Geoscience Research programme under award number (LANLE3W1). Secondary support is from the Consortium Advancing Technology for Assessment of Lost Oil & Gas, funded by US Department of Energy, Office of Fossil Energy and Carbon Management, Office of Resource Sustainability, Methane Mitigation Technologies Division’s, Undocumented Orphan Wells Program. This paper has been co-authored by UT-Battelle, LLC under contract no. DE-AC05-00OR22725 with the US Department of Energy. J.E.S. thanks A. Jaegle and J. Carreira for their useful suggestions. Finally, we are grateful to the developers of the many software packages used throughout this project including, but not limited, to PyTorch, Numpy, Vedo, Matplotlib and PyTorch Lightning.

FundersFunder number
Center for Non-Linear Studies
U.S. Department of Energy
Office of Science
Basic Energy SciencesLANLE3W1
Laboratory Directed Research and Development
UT-BattelleDE-AC05-00OR22725
Office of Fossil Energy and Carbon Management
Office of Resource Sustainability

    Fingerprint

    Dive into the research topics of 'Development of the Senseiver for efficient field reconstruction from sparse observations'. Together they form a unique fingerprint.

    Cite this