U.S. Department of Energy, Office of Electricity

Faster from lab to market for long-duration energy storage

ROVI brings six national laboratories together to predict how energy storage systems age. Adaptive AI/ML models learn from lab tests and simulations, then get validated against field-deployed systems, so new technologies need far less real-time testing.

Open the ROVI DataHub Read our mission
National laboratories
6
Researchers
24
Open-source repositories
10

Predicting battery lifetime early

Capacity retained vs. charge cycles

Illustrative chart: measured capacity from the first 800 cycles of testing, and an AI model forecasting the rest of the curve with an uncertainty band, reaching the 80% end-of-life threshold near 3,100 cycles.
  • Lab measurements
  • AI/ML forecast
  • Uncertainty
Illustrative example of early lifetime prediction, not real test data.

Six national laboratories, one initiative

Testing, modeling and data infrastructure expertise from across the DOE lab system.

Explore ROVI

DataHub

One secure home for energy storage data

Lab tests, Li-ion field streams and flow battery data, consolidated for evaluation and machine learning.

See the architecture Team

24 researchers across DOE and six labs

Meet the team Mission

Lab to market, faster

Why ROVI exists and how it shortens validation.

Read more Publications 12

Papers, reports, datasets and software records since 2021.

Browse Software 10

Open-source repositories on GitHub.

View code Latest talk

ROVI DataHub: unifying energy storage data

May 2026, ORNL

All talks Industry 3

Long-duration storage demonstration partners.

See partners Resources

Platform, dashboards, API and toolkit

Open tools
Our mission

Greatly reduce the time required for emerging energy storage technologies to go from lab to market.

DOE's Rapid Operational Validation Initiative addresses critical gaps in the data needed to evaluate the long-term performance of energy storage technologies, and accelerates the pace of technology development.

Many innovative storage technologies are promising candidates to meet cost and performance targets such as DOE's Long Duration Storage Shot, and to reach the widespread deployment a reliable, clean and affordable grid will need. ROVI develops tools that accelerate the testing and validation each of them must pass to succeed commercially.

Using artificial intelligence and machine learning on large sets of storage performance data at different scales, ROVI generates lifetime performance predictions for new technologies with minimal real-time testing. Models are validated with data from field systems supported by DOE's Office of Electricity (OE) and Office of Clean Energy Demonstrations (OCED).

DOE program page

How ROVI shortens validation

1

Test and simulate

Laboratory cell testing and model simulations across chemistries produce consistent, well-described data sets.

2

Learn and predict

Adaptive AI/ML models forecast degradation and operational lifetime from early-life data.

3

Validate in the field

Predictions are checked against field-deployed systems supported by DOE OE and OCED demonstrations.

DataHub development

One secure home for energy storage data

The ROVI team built a flexible, secure data architecture that ingests complex data streams and supplies consolidated data sets for project evaluations and advanced machine learning.

Lab cell testsCAMP, accelerated testing Li-ion system dataField streams, Sandia Flow battery dataRedox flow, PNNL DataHubhosted at ORNL AI/ML modelsHealth and lifetime DashboardsLive monitoring Data APIPython via rovidh

  • Machine learning models

    Predict battery degradation and performance across operating conditions and environmental factors.

  • Data validation framework

    Protocols that keep data quality and consistency across laboratory and field deployment sites.

  • Real-time analytics

    Live monitoring and analysis for continuous assessment of system performance and health.

  • Collaborative research

    Shared data access and coordinated research across DOE facilities.

Team

The people behind ROVI

Researchers and program leads across the Department of Energy and its national laboratories. Names link to profile pages where available.

Alumni

Former ROVI researchers and program contributors.

Publications

Research and reports

Featured paper

Multivariate prognosis of battery advanced state of health via transformers

Noah H. Paulson, Joseph J. Kubal, Susan J. Babinec. Cell Reports Physical Science 5, 101928 (2024).

Simultaneous prognosis of 28 battery health descriptors across six lithium-ion cathode chemistries, including an unprecedented mean absolute error of 19 cycles for end-of-life on a fast-charging LFP dataset.

Read the paper (PDF) doi:10.1016/j.xcrp.2024.101928
2024
2023
  • Lithium-ion ROVI Requirements Overview U.S. Department of Energy, Office of Electricity (2023). ReportPDF
  • Electrical Energy Storage Data Submission Guidelines Sandia National Laboratories. SAND2023-12079 (2023). ReportPDF
  • EZBattery hybrid analytical flow-battery model (part III) Yunxiang Chen, et al. Journal of Power Sources 578, 233210 (2023). doi:10.1016/j.jpowsour.2023.233210
  • Dataset of NMC battery tests from CAMP, 2023 release Logan Ward, Joseph Kubal, Susan J. Babinec, Wenquan Lu, Allison Dunlop, Steve Trask, et al. Datasetdoi:10.18126/p0xr-uqfp
2022
  • Feature engineering for machine learning enabled early prediction of battery lifetime Noah H. Paulson, Joseph Kubal, Logan Ward, Saurabh Saxena, Wenquan Lu, Susan J. Babinec. Journal of Power Sources 527, 231127 (2022). doi:10.1016/j.jpowsour.2022.231127
  • Principles of the battery data genome Logan Ward, Susan Babinec, Eric J. Dufek, David A. Howey, Venkatasubramanian Viswanathan, et al. Joule 6, 2253–2271 (2022). doi:10.1016/j.joule.2022.08.008
  • Machine-learning assisted identification of accurate battery lifetime models with uncertainty P. Gasper, et al. Journal of The Electrochemical Society 169, 080518 (2022). NLR record
2021
Docs
  • Battery Data Toolkit (battdat) Documentation for ROVI's battery data library. SoftwareDocumentation
Software

Open-source tools

ROVI develops open-source tools for battery data handling, state-of-health estimation, and prognostics.

ROVI-org on GitHub

Browse repositories, open issues, and contribute.

github.com/ROVI-org

battery-data-toolkit

Utilities for reading and manipulating battery testing data.

PythonApache-2.0
Repository Docs

auto-soh (Moirae)

Fit state of health (SOH) to batteries using state estimation techniques.

PythonMIT
Repository Docs

rovidh

ROVI DataHub utility for interacting with the DataHub from Python.

PythonMIT
Repository

All repositories

RepositoryDescriptionLanguageLicense
rovidhROVI DataHub utilityPythonMIT
auto-sohFit state of health to batteries using state estimationPythonMIT
battery-data-toolkitRead and manipulate battery testing dataPythonApache-2.0
services-demonstrationPipeline from data upload to a system health dashboardPythonMIT
collecting-battery-datasetsDownload and standardize data used by ROVI modeling teamsPythonMIT
model-to-prognosisExample pipeline: model → synthetic data → AI → prognosisPythonMIT
storage-requirement-estimationEstimate ROVI data storage requirementsPythonMIT
theveninEquivalent circuit models in Python (fork of NatLabRockies/thevenin)PythonBSD-3-Clause
progpyNASA prognostics packages for remaining-useful-life modeling (fork of nasa/progpy)PythonNASA OSA
rovi-org.github.ioSource for this websiteHTMLMIT
Presentations

Talks, webinars and briefings

  1. May 2026ORNL

    ROVI DataHub: unifying energy storage data

    Demonstration of the ROVI DataHub to the DOE Office of Electricity Energy Storage program, combining Li-ion system streams from Sandia with redox flow battery datasets from PNNL.

    Watch video ORNL news story
  2. 2023DOE Office of Electricity

    Rapid Operational Validation Initiative (ROVI) Overview

    Presented by Vinod Siberry (DOE OE) and Vincent Sprenkle (PNNL), with panelists Yuliya Preger (SNL), Srikanth Allu (ORNL), Noah Paulson (ANL), Jie Bao (PNNL), and Kevin Gering (INL) covering data requirements, infrastructure, AI/ML, and accelerated testing.

    Slides (PDF)
  3. 2023DOE Office of Electricity

    Lithium-ion ROVI Requirements Overview

    Data requirements and reference performance tests for Li-ion systems contributing to ROVI.

    Slides (PDF)
  4. 2021Industry Roundtable

    Rapid Operational Validation Initiative: Industry Roundtable

    Launch discussion with industry and national laboratory participants on validating new storage technologies faster than conventional testing allows.

    Slides (PDF)
Industry collaborations

Validated against real deployments

ROVI works with technology developers, utilities, and industry groups to shape data requirements and validate models against field-deployed systems.

Long-duration storage demonstration partners

Solid-flow battery

CMBlu Energy

Long-duration solid-flow battery demonstration with Argonne and Idaho National Laboratories.

Lab partners: ANL, INL
cmblu.com
Zinc battery

e-Zinc

Multi-day zinc battery demonstration as an alternative to diesel backup during utility power shutoffs.

Lab partner: SNL
e-zinc.ca
Vanadium flow

Invinity Energy Systems

24-hour vanadium flow battery demonstration.

Lab partner: PNNL
invinity.com

These projects were selected under the DOE Long-Duration Energy Storage Demonstrations lab call, a source of field-deployed systems for validating ROVI models.

Electric Power Research Institute (EPRI)

Industry Roundtable participant; EPRI energy storage data guidelines inform ROVI data requirements.

epri.com

New Energy Risk

Industry Roundtable participant bringing a performance-risk and insurance perspective.

newenergyrisk.com

Battery Council International

BCI / Cleantech Strategies Flow Battery Working Group engagement on flow battery data needs.

batterycouncil.org

Contribute data to ROVI

Project teams can complete the workplan template on the OE ROVI page and send it to ESGC@hq.doe.gov.

Email ESGC@hq.doe.gov Get the workplan template