Data Science and Systems Integration
Research and Development Projects
Data Science & Systems Integration (DSSI) continuously improves and expands efforts across the facility. Here are a few key projects that DSSI is currently engaged in.
ILLUMINE Project
The Intelligent Learning for Light Source and Neutron Source User Measurements Including Navigation and Experiment Steering (ILLUMINE) project is a collaborative effort that spans all light and neutron sources in the DOE complex. The goals of the project are to:
- Develop multi-facility framework based on Bluesky to address data volume and complexity at X-ray/neutron sources
- Enable rapid data analysis and autonomous experiment steering
- Leverage real-time compression, machine learning, and decision support techniques to optimize data collection
Flyscanning
Collecting synchrotron data "on the fly" rather than point-by-point makes use of specialized equipment to dramatically improve throughput. Instead of waiting for conditions to stabilize at each point along a scan, data can be collected by taking snapshots as changes occur continuously over a wide range of interest. In order to support this capability at facility scale, DSSI has identified standard hardware and software approaches which meet the needs of multiple techniques. Broadly speaking, these flyscanning approaches include:
- Measurements and position data are collected synchronously. (time-based)
- Measurements are collected at specific pre-defined positions. (position-based)
- Measurement and position data are collected independently, using time stamps and interpolation to co-bin streams. (asynchronous)
Robotics
DSSI teams have built open-source robotics solutions for self-driving experiments and continue to design and develop contemporary robotic systems to accelerate the scientific workflows at NSLS-II. They leverage the community supported Robot Operating System (ROS) for control with the Bluesky Project for higher level orchestration to ensure broad applicability of the projects around NSLS-II and beyond. Researchers also endow robotic systems with computer vision environment recognition and artificial intelligence for experiment decision making.
