EDS Projects
LCLS leads and participates in numerous collaborative research projects spanning artificial intelligence, high-performance computing, and advanced data systems. These initiatives bring together experts from national laboratories, universities, and research institutions to tackle the unprecedented data challenges posed by next-generation light sources.
These partnerships drive innovation in real-time data processing, experiment automation, and scientific discovery through cutting-edge technologies. By leveraging complementary expertise across the DOE complex and beyond, these projects create transformative capabilities that enhance the scientific impact of LCLS experiments and benefit the broader scientific community.
Dive into our Research
ILLUMINE
(LCLS, SSRL, ALS, APS, NSLS-II, SNS/HFIR)
Integrated artificial intelligence methods for accelerated multimodal analysis of chemical systems.
Key Objectives:
AI-accelerated analysis of chemical systems
Multi-facility data integration
Automated experimental steering
ExaFEL
(LCLS/NERSC)
Exascale computing for free-electron laser data analysis enabling real-time processing of crystallography data.
Key Objectives:
High-performance computing for FEL data
Real-time feedback for experiments
Scalable analysis algorithms
AISDC
(LCLS/ANL)
AI-enabled Scientific Data Center that accelerates scientific discovery through machine learning and advanced data techniques.
Key Objectives:
AI-driven experimental control
Automated data processing pipelines
Cross-facility data integration
Diaspora
(ANL led, LCLS is a partner)
Resilience-enabling services for science from HPC to edge.
Key Objectives:
Create a hierarchical event fabric
Develop resilience services
Evaluate new capabilities in scientific applications
LCLStream
(LCLS/Oak Ridge)
Real-time streaming data analysis framework for LCLS experimental data, enabling on-the-fly processing and visualization.
Key Objectives:
Streaming data processing
Real-time visualization
Reduced data latency
SparkPix-RT
(LCLS/Argonne)
Real-time pixel detector data processing using distributed computing frameworks for high-throughput analysis.
Key Objectives:
Parallel processing of detector data
High-throughput analysis
Scalable framework for megapixel detectors
LLAna
(LCLS/LBNL)
The LLAna project enhanced LCLS data analysis by improving HDF5 interoperability, optimizing workflows, and scaling Jupyter for HPC.
Key Objectives:
Support HDF5 read-while-write for LCLS-II
Scale Jupyter for large-scale data processing
Optimize HPC workflows for I/O-intensive tasks
AUREIS
(SLAC, ANL, BNL, LBNL, FNAL, LLNL, Stanford University)
The AUREIS project develops adaptive, ultra-fast sensing technologies using AI/ML, advanced ASICs, and wide-bandgap materials for scientific imaging and microelectronics research.
Key Objectives:
On chip AI/ML-driven workflows for dynamic experiment control.
Energy-efficient edge computing systems and adaptive ASICs
Ultra-wide-bandgap (UWBG) materials for efficient multi-energy detection.
Legion & SpiniFEL
(SLAC CS)
Legion is a data-centric parallel programming system for writing portable high performance programs targeted at distributed heterogeneous architectures.
Key Objectives:
Develop an exascale single-particle imaging code (SpiniFEL)
Evaluate tradeoffs between MPI and Pygion
MLCV Projects
Machine Learning and Computer Vision projects enhancing LCLS capabilities:
- Computer vision for automated sample analysis
- ML-based experimental optimization
- Automated anomaly detection in experiments
