{"resourceId":"doe-synaps-i-intelligent-xray-imaging-nsls2-20261104","versions":[{"version":"external-789b4312e72fb4b11f52795054760bd6e05fbfabd7ac2995031f355a579a9ac4","resource":{"id":"doe-synaps-i-intelligent-xray-imaging-nsls2-20261104","title":"SYNAPS-I: Intelligent Computing Transforms X-ray Imaging, from Data Collection to Discovery","organization":"Brookhaven National Laboratory (NSLS-II), with Berkeley Lab ALS, Argonne, Oak Ridge and SLAC; DOE Genesis Mission ModCon seed project","sector":"National laboratory user facilities (AI-driven experimental workflows)","geography":"United States (multi-laboratory DOE)","publishedAt":"4 November 2026 (BNL newsroom)","publicationDate":"2026-11-04","eventDate":null,"sourceName":"Brookhaven National Laboratory Newsroom","sourceLabel":"BNL Newsroom, \"SYNAPS-I: Intelligent Computing Transforms X-ray Imaging, from Data Collection to Discovery,\" 4 Nov 2026","sourceUrl":"https://www.bnl.gov/newsroom/news.php?a=223173","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["infrastructure","developers-agents","operating-model"],"finding":"Brookhaven reports SYNAPS-I demonstrations at NSLS-II. One automated experiment analyzed images as acquired and autonomously selected scientifically interesting regions for high-resolution measurement without human intervention. A foundation model trained at Argonne replaced iterative ptychographic reconstruction and, integrated with NVIDIA Holoscan, ran real-time reconstruction at 200 images per second. A model trained at NSLS-II reconstructed a microchip from nearly one tenth of the usual data, and fine-tuned Meta SAM3 segmented the 3D volumes. Next steps are real-time tomography and more general autonomous workflows.","sledRelevance":"Relevant to research universities that use DOE light sources and to campus imaging cores: AI is shifting beamlines from collect-then-analyze to data-guided experiments, with implications for user data pipelines and GPU needs. Limited direct relevance outside imaging-intensive research.","evidence":"Laboratory press release describing demonstrations; no error rates, validation against conventional reconstruction, or beam-time savings quantified beyond the 200 images per second and roughly one-tenth data claims. NVIDIA and Meta tools named. Classified as a self-reported institutional claim.","architectureImplications":"Edge streaming reconstruction (NVIDIA Holoscan) at the beamline, foundation models trained at partner labs, integration with facility data infrastructure; moves compute into the experiment loop.","governanceImplications":"Autonomous selection of measurement regions shifts decisions from users to models; facilities need validation and user consent norms. Not discussed.","securityPrivacyImplications":"Microelectronics samples may be proprietary or export-controlled; data handling not described.","caveats":"Demonstrations only; quantitative claims self-reported; sample types (thin, flat microchips) favour the reduced-data approach; generalization unproven.","streamIds":["research"],"roles":{"sales":"Facility framing: DOE labs are demonstrating data-guided beamline experiments and real-time reconstruction at 200 images per second. That supports bounded conversations with university imaging cores and research computing about streaming analysis and GPU-at-the-edge needs. Interpretation: the release is a self-report without validation data, so it cannot support claims of guaranteed beam-time or cost savings.","engineering":"Fit is imaging workflows with heavy reconstruction loads. Architecture places streaming inference (Holoscan) near instruments, foundation models for reconstruction and segmentation, and links to facility data systems. Prerequisites are GPUs at the edge, high-bandwidth data paths and trained models. Constraints: results shown on thin, flat microchips. A proof of value should compare learned and iterative reconstruction quality and data reduction on local samples.","delivery":"Delivery work is adding streaming reconstruction and automated region selection to an imaging workflow, with model validation and user-facing controls. Ownership sits with the facility data science group and beamline or core staff. Governance checkpoints include validation before scientific use and data-handling rules for proprietary samples. Acceptance criteria are proposals such as measured reconstruction fidelity versus baseline; risks include model drift across sample types."},"roleDetails":{"sales":{"customerProblem":"Imaging experiments produce data faster than staff can reconstruct and analyze, wasting scarce instrument time.","stakeholders":["Imaging core director","Research computing director","Beamline users and PIs","Facility data science lead"],"discoveryQuestions":["Where are reconstruction bottlenecks in your imaging workflows?","Is GPU capacity available near instruments?","How would learned reconstructions be validated?","Who decides when models select measurement regions?"],"valueHypothesis":"Streaming learned reconstruction may cut data collection and analysis time for suitable sample types, as BNL reports for microchips.","potentialEngagement":"A workflow assessment for one imaging instrument covering streaming reconstruction, data paths and validation.","unsupportedClaims":["Beam-time or cost savings","Accuracy across sample types","Autonomous experiments without human oversight"]},"engineering":{"fit":"Strong for ptychography and tomography with reconstruction bottlenecks.","architectureIntegration":"Edge GPU streaming inference, foundation reconstruction models, segmentation models, facility data infrastructure.","prerequisites":["Edge GPUs","High-bandwidth data paths","Trained models and validation datasets"],"constraints":["Demonstrations on thin, flat samples","Self-reported metrics","Export control for some samples"],"security":"Control access to sample data; separate proprietary user data; log model versions used per experiment.","proofOfValue":["Compare learned vs iterative reconstruction fidelity","Measure data reduction on local samples","Audit autonomous region selection against expert choices"]},"delivery":{"work":"Integrate streaming reconstruction; Validate models and expose user controls","dependencies":["Facility data infrastructure","GPU procurement","Model training partners"],"ownership":"Facility data science group with beamline or core staff.","skillsAdoption":"Staff need ML operations and validation skills; users need guidance on model-guided experiments.","governanceCheckpoints":["Model validation before scientific use","User consent for autonomous region selection","Data-handling review for proprietary samples"],"acceptanceCriteria":["Proposed: reconstruction fidelity documented versus baseline","Proposed: data reduction measured per sample type","Proposed: model version recorded for every dataset"],"risks":["Model drift across sample types","Missed features in unselected regions","Vendor platform dependence"]}},"retrievedAt":"2026-11-04T22:01:17-06:00","enrichedAt":"2026-11-04T22:01:19-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Facility staff need ML engineering and data infrastructure skills; collaboration across labs is emphasized.","procurementImplications":"Named NVIDIA Holoscan and Meta SAM3; facilities should plan for edge GPU purchases and open-weight model use.","operatingModelImplications":"Multi-lab collaboration model with facility data science teams integrating tools for users.","sourceVerification":{"openedUrl":"https://www.bnl.gov/newsroom/news.php?a=223173","referenceExcerpt":"Working with NVIDIA, the team integrated the workflow with the NVIDIA Holoscan platform and demonstrated real-time ptychographic reconstruction at 200 images per second.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}