Speaker
Description
We present the current status of the data acquisition (DAQ) project at the Canadian Light Source, designed to provide a generic, scalable solution for beamline instrumentation, minimizing the need for bespoke implementations. Built on the bluesky ecosystem, the framework orchestrates data collection from EPICS-controlled devices and PandABox hardware, streaming structured documents through a persistent NATS Jetstream message queue with beamline-specific subjects.
Data is ingested via Tiled into a PostgreSQL backend, enabling efficient indexing, querying, and future extensibility. The system leverages pixi for reproducible beamline-specific environments and CI/CD workflows, ensuring maintainability and deployment consistency.
A key integration challenge has been the reliable execution of Python-based DAQ processes within a systemd-managed environment, which we address through operating system package built in CI/CD-pipelines. While the backend is production-ready, frontend development is ongoing, with a full rollout planned for early 2026.
This talk will outline the architecture, integration lessons, and future roadmap, including live visualization, metadata enrichment, and broader device compatibility.
