DFS Recipes

Use recipes when you already know the job you need DFS to complete. Each recipe combines DFS Lite connection and mapping steps with DFS Pro governance, fusion, review, or reporting steps where they are needed.

Recipe decision flow

Choose a recipe

Goal	Recipe	Use when
Bring facility signals into a twin	Connect BMS Data to a Facility Twin	A building, plant, or campus team needs BMS, meter, alarm, or equipment signals available in FactVerse.
Prepare signal history for equipment analysis	Prepare Signal History for Predictive Maintenance	Maintenance and reliability teams need clean time-series history before using predictive maintenance workflows.
Package reviewed data for Agent workflows	Create an AI Agent-Ready Dataset	An AI Agent workflow needs governed source data, stable IDs, metadata, and reviewer context.
Combine operations records and sensor data	Fuse Inspection, Work Order, and Sensor Data	A workflow needs one operational view across inspections, work orders, alarms, and live or historical signals.
Recover from rejected source rows	Fix Rejected Rows and Reprocess	A sync or fusion job rejected rows that need source correction, manual resolution, or controlled reprocessing.

Recipe pattern

Most DFS recipes follow the same operating loop:

Define target workflow
  -> connect source
  -> preview and map data
  -> sync and check quality
  -> promote to governed dataset when reuse is needed
  -> review uncertain outputs
  -> publish to the consuming FactVerse workflow

Before you start

Prepare these inputs before following a recipe:

• tenant and project context;
• source owner and reviewer;
• source connection details;
• target asset, point, dataset, or workflow identity;
• expected units, ranges, timestamps, and update frequency;
• acceptance criteria for sync, data quality, and reviewer handover.

Expected outputs

Output	Use
Source and target boundary	Keeps the recipe tied to one operating problem and one accountable owner.
Connector and mapping evidence	Shows how source data enters DFS and which target identity it supports.
Quality and sync note	Explains whether the source is reliable enough for the chosen workflow.
Dataset, fusion, or review output	Captures reusable governed data when the recipe needs DFS Pro.
Handoff record	Gives the consuming team source context, known limits, and reviewer decisions.

Recipe completion rule

A recipe is complete when the consuming team can repeat the data path without asking the author to explain hidden assumptions. Capture the source owner, target identity, sync or dataset version, quality status, review decision, and downstream owner. For Agent workflows, include the allowed answer type and evidence boundary. For facility or maintenance workflows, include the site, asset group, signal window, and known source limits.

Related documentation

Page	Use
Getting Started with DFS	Create a first connector and verify a small end-to-end data loop.
Prepare DFS Data for AI Agent Workflows	Use DFS Lite and DFS Pro together before an Agent workflow reads operational data.
DFS Lite	Connect, preview, map, sync, and check source data.
DFS Pro	Create governed datasets, fusion tasks, review queues, pipelines, and BI reports.
DFS Reference	Check connector types, mapping fields, permissions, and API surfaces.