Skip to content
DraconiticResearch
Search Sources API
186,123 studies31 sourcesUpdated 24 min ago
PreprintCC BYChemRxiv

From Automated Beamlines to Autonomous Science

  • Chenhui Zhu1,2
  • Marcus M. Noack3,2
  • Wiebke Koepp1,2
  • Thomas Ferron1,2
  • Oleg Kostko1,2
  • Damon English1,2
  • Jeff Dickert4
  • Chris Jozwiak1,2
  • Jinghua Guo1,2
  • Dilworth Y. Parkinson1,2
4 affiliations
  1. 1Advanced Light Source
  2. 2Lawrence Berkeley National Laboratory
  3. 3Applied Mathematics and Computational Research Division
  4. 4Molecular Biophysics & Integrated Bioimaging, Lawrence Berkeley National Laboratory
Open at sourcePDFDOICite

Abstract

Autonomous experimentation is emerging across materials synthesis, synchrotron and neutron characterization, real-time analysis, and scientific decision-making. Yet increasingly capable artificial intelligence does not by itself create autonomous science. Scientific user facilities must also provide machine-operable experiments, machine-readable and interoperable data and interfaces, persistent provenance across the experimental lifecycle, scientific capabilities that can be discovered and orchestrated across instruments and facilities, and access models compatible with iterative experimental campaigns. We argue that these infrastructure requirements should be treated as a coupled scientific design problem. Their value is not limited to higher throughput: reducing the latency between observation, interpretation, decision, and the next measurement can make metastable states, competing phases / phase coexistence, degradation, failure pathways, phase boundaries, and other difficult-to-access phenomena experimentally observable. Drawing on recent demonstrations in autonomous laboratories and advanced characterization, along with discussions from the 2026 Advanced Light Source Automation and Self-Driving Labs Workshop, we outline a practical progression from automated instruments to distributed autonomous discovery ecosystems that link synthesis, synchrotron and neutron characterization, computing, theory, and human scientific judgment.

Related

  1. An Automated Scholarly Literature Collection Workflow Using Official Services for AI for Science
    Yuxing Gong, Zechen Li, Hailun Xu +7
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  2. Bridging the Literature-to-Laboratory Gap with Language Models and Active Learning for Reaction Feasibility Prediction and Autonomous Exploration
    Haowen Zhong, Yi Yang, Jiekai Jia +7
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  3. Autonomous Experimentation Enables Rapid Optimization of Photochemical Continuous-Flow Synthesis of Diazaxanthilidene Fluorophores
    Joao Luiz P. de Albuquerque, Scott Laverty, Mao Niimura +3
    ChemRxivPreprintCC BYFull textGeneral chemistryChemRxiv
  4. From Signals to Measurands: A Measurement-Science Roadmap for Reproducible Analytical Biochemistry
    Santiago Schnell
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  5. Spotlighting rising researcher Alaa Darwesh: engineering smart nanoscaffolds for ocular therapy: merging cocrystal science with electrospun nanofibers
    Alaa Y. Darwesh
    AAPS OpenPeer-reviewedCC BYFull textMaterials & polymersDOAJ

Cite

BibTeXRIS
BibTeX
@misc{zhu2026from,
  title = {{From Automated Beamlines to Autonomous Science}},
  author = {Chenhui Zhu and Marcus M. Noack and Wiebke Koepp and Thomas Ferron and Oleg Kostko and Damon English and Jeff Dickert and Chris Jozwiak and Jinghua Guo and Dilworth Y. Parkinson},
  howpublished = {ChemRxiv},
  year = {2026},
  month = sep,
  doi = {10.26434/chemrxiv.15009126/v1},
  url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15009126/v1},
  note = {CC BY},
}
RIS
TY  - UNPB
TI  - From Automated Beamlines to Autonomous Science
AU  - Chenhui Zhu
AU  - Marcus M. Noack
AU  - Wiebke Koepp
AU  - Thomas Ferron
AU  - Oleg Kostko
AU  - Damon English
AU  - Jeff Dickert
AU  - Chris Jozwiak
AU  - Jinghua Guo
AU  - Dilworth Y. Parkinson
PY  - 2026
DA  - 2026/09/21
PB  - ChemRxiv
DO  - 10.26434/chemrxiv.15009126/v1
UR  - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15009126/v1
AB  - Autonomous experimentation is emerging across materials synthesis, synchrotron and neutron characterization, real-time analysis, and scientific decision-making. Yet increasingly capable artificial intelligence does not by itself create autonomous science. Scientific user facilities must also provide machine-operable experiments, machine-readable and interoperable data and interfaces, persistent provenance across the experimental lifecycle, scientific capabilities that can be discovered and orchestrated across instruments and facilities, and access models compatible with iterative experimental campaigns. We argue that these infrastructure requirements should be treated as a coupled scientific design problem. Their value is not limited to higher throughput: reducing the latency between observation, interpretation, decision, and the next measurement can make metastable states, competing phases / phase coexistence, degradation, failure pathways, phase boundaries, and other difficult-to-access phenomena experimentally observable. Drawing on recent demonstrations in autonomous laboratories and advanced characterization, along with discussions from the 2026 Advanced Light Source Automation and Self-Driving Labs Workshop, we outline a practical progression from automated instruments to distributed autonomous discovery ecosystems that link synthesis, synchrotron and neutron characterization, computing, theory, and human scientific judgment.
ER  -