We are defining a reusable tag namespace for a liquid-cooled LPBF copper cold-plate validation rig. The controller needs to publish inlet/outlet temperature, flow rate, pressure drop, leak status, pump state, and test-run metadata so that results remain traceable across geometry revisions.
A useful reference for separating simulated inputs from measured acceptance criteria is this thermal design and validation workflow.
For a groov EPIC/RIO system using Sparkplug B, I am considering:
- one stable device identity for the test cell;
- metrics for sensors and calculated delta-T / thermal resistance;
- run ID, part revision, material, and fixture revision as properties rather than dynamic metric names;
- birth/death certificates for cell availability, but explicit run-state metrics for test lifecycle.
Would you keep all runs under one Edge Node and Device, or create a device identity per fixture? Also, what has worked best for preserving engineering context without causing metric churn in the UNS?