DFS Workflows

Use workflow guides when a data path crosses DFS Lite, DFS Pro, MDM, and a consuming FactVerse application. A workflow starts from the business outcome, not from a DFS feature name. It shows which source systems are needed, when data can stay in DFS Lite, when it should be promoted into DFS Pro, when MDM identity governance is required, and what evidence the downstream owner needs before using the output.

DFS workflows are most useful when facility, maintenance, data, and application teams share responsibility for the same operational dataset. The workflow page keeps the teams aligned on source ownership, mapping decisions, review boundaries, dataset versioning, and handoff evidence.

Before you start

Prepare the downstream owner, source owner, target site or asset scope, expected fields, update frequency, permission boundary, reviewer, and acceptance criteria before choosing a workflow. If the same data will be reused by more than one team or application, plan for DFS Pro governance and MDM identity review before the first production handoff.

Choose by outcome

Target outcome	Recommended path	Use when	Key evidence
Facility signals available in an operational digital twin	BMS/source data to facility digital twin	A building, plant, data center, or campus team needs BMS, meter, alarm, or equipment signals mapped to FactVerse assets and points.	Source owner, connector, point mapping, unit normalization, sync history, quality note, consuming twin scope.
Signal history ready for predictive maintenance	Predictive maintenance signal history	Maintenance teams need reliable time-series history before using FactVerse AI Agent predictive maintenance workflows.	Asset identity, signal window, sampling rules, missing-value handling, quality review, model-readiness note.
Reviewed data package for Agent workflows	AI Agent-ready dataset	An Agent workflow needs governed source data, stable IDs, metadata, and reviewer context before it can answer or trigger tasks.	Allowed answer type, source boundary, dataset version, reviewer decision, evidence contract.
Stable identity across systems	MDM identity resolution and alias governance	Different systems describe the same asset, location, device, document, or work order with different IDs or names.	Master entity, alias record, merge/split/re-point decision, steward queue result, audit note.
One view across inspections, work orders, alarms, and sensor history	Fault, alert, and event fusion	Operations teams need to correlate issue records, alarm events, maintenance actions, and signal history.	Fusion task, conflict handling, rejected-row path, review queue decision, downstream owner.
Governed data available for BI or reporting	DFS Pro dataset to BI/reporting	A recurring report or dashboard needs a governed dataset with versioning, lineage, and access control.	Dataset version, fields, refresh cadence, permission scope, report owner.
Lite connector output promoted into a reusable governed asset	Lite-to-Pro governed operational pipeline	A source connection starts as a local integration, then needs repeatable governance, fusion, or reuse.	Connector, mapping, quality gate, dataset promotion, lineage, handoff record.

Selection flow

Operating model

Stage	DFS area	What to decide	Output
Define the target	Workflow owner	Which operational decision, page, Agent workflow, report, or simulation will use the data.	Evidence contract and acceptance criteria.
Connect and map	DFS Lite	Which source system, connector, fields, units, timestamps, and target IDs are authoritative enough to start.	Connector, preview, mapping, sync, and quality record.
Promote when reusable	DFS Pro	Whether the data needs versioning, lineage, access control, fusion, review, or BI reuse.	Governed dataset, fusion task, review queue item, or reporting dataset.
Resolve identity	DFS MDM	Whether aliases, duplicate entities, or cross-system naming differences can affect the result.	Master entity, alias, merge/split/re-point decision, steward audit note.
Hand off	Consuming workflow owner	Whether the output is current, complete, reviewed, and allowed for the target workflow.	Handoff record with owner, scope, limits, and validation status.

BMS/source data to facility digital twin

Use this path when facility or campus signals need to appear in an operational digital twin. Start with the Connect BMS Data to a Facility Twin recipe, then add governance only when the same points will be reused across sites, reports, Agent workflows, or maintenance processes.

Step	Page to use	Decision
Connect source	DFS Lite connectors	Select the BMS, meter, alarm, historian, or equipment source and confirm owner access.
Map fields	DFS Lite mappings	Bind source points to target asset, location, point name, unit, timestamp, and update frequency.
Check quality	DFS Lite data quality	Confirm missing values, stale points, rejected rows, and unit consistency before handoff.
Promote if needed	DFS Pro datasets	Create a governed dataset when multiple teams or workflows will reuse the same facility data.

The handoff should name the site, equipment scope, point list, source freshness, known gaps, and the consuming twin or application page.

Predictive maintenance signal history

Use this path when a maintenance or reliability team needs signal history before running a predictive maintenance workflow. Start with the Prepare Signal History for Predictive Maintenance recipe. Use DFS Pro when the same history becomes a maintained feature dataset, and use MDM when asset names differ across historian, CMMS, inspection, and work-order systems.

Step	Page to use	Decision
Define asset and signal scope	Mapping fields reference	Confirm asset ID, signal name, timestamp, unit, sampling window, and accepted time range.
Prepare source history	DFS Lite sync history	Check whether enough clean history exists for the target equipment group.
Govern the dataset	DFS Pro dataset lifecycle	Version the signal history when it feeds recurring model validation or review.
Resolve asset identity	MDM entity resolution tasks	Review duplicate or conflicting asset identities before the dataset is used.

The handoff should include asset group, signal list, history window, sampling treatment, quality issues, and the reviewer who accepted the dataset for predictive maintenance use.

AI Agent-ready dataset

Use this path when a FactVerse AI Agent workflow needs governed data before it can answer questions, produce recommendations, or prepare an operating task. The workflow guide Prepare DFS Data for AI Agent Workflows covers the full sequence.

Step	Page to use	Decision
Define evidence boundary	Prepare DFS Data for AI Agent Workflows	Specify the allowed answer type, source systems, target identity, freshness, and reviewer boundary.
Build source and dataset path	Create an AI Agent-Ready Dataset	Combine Lite mapping, Pro dataset governance, and reviewer notes.
Review uncertain outputs	DFS Pro review queue	Route rejected, uncertain, or conflicting rows before Agent use.
Record permissions	DFS permissions reference	Confirm who can create, review, publish, and consume the dataset.

The handoff should state what the Agent can use the dataset for, what it must not infer from missing data, and which reviewer accepted the evidence boundary.

MDM identity resolution and alias governance

Use this path when the same real-world object has different identifiers across BMS, CMMS, ERP, inspection, document, or model sources. Start with DFS MDM and use cross-source aliases to preserve traceability instead of forcing every source to use the same native identifier.

Step	Page to use	Decision
Define master entity	Master entities	Decide the entity type, canonical label, lifecycle state, and responsible steward.
Resolve duplicates	Entity resolution tasks	Review suggested matches, confidence, evidence, and rejection reasons.
Govern aliases	Cross-source aliases	Keep source-specific IDs visible while linking them to the master entity.
Correct identity decisions	Merge, split, and re-point	Apply controlled corrections when the identity graph changes.

The handoff should include master entity ID, alias sources, steward decision, affected datasets, and whether downstream mappings need reprocessing.

Fault, alert, and event fusion

Use this path when operations teams need one reviewed view across inspection records, work orders, alarms, incidents, and sensor history. Start with the Fuse Inspection, Work Order, and Sensor Data recipe, then use fault event fusion when event identity and correlation need governance.

Step	Page to use	Decision
Gather source records	Connector types reference	Confirm which operational systems provide inspection, alarm, sensor, work-order, or event records.
Create fusion task	DFS Pro fusion tasks	Define join keys, time windows, conflict rules, and review routing.
Review outputs	DFS Pro review queue	Resolve uncertain matches, rejected rows, duplicate events, and reviewer notes.
Recover bad rows	Fix Rejected Rows and Reprocess	Correct source, mapping, or identity issues and rerun the controlled path.

The handoff should name the event scope, time window, source systems, match rules, unresolved conflicts, and the team that owns the downstream response.

DFS Pro dataset to BI/reporting

Use this path when a report, dashboard, or recurring analysis needs a stable governed dataset. DFS Pro should provide versioning, lineage, refresh cadence, permissions, and audit history before the BI owner treats the output as a reporting source.

Step	Page to use	Decision
Create governed dataset	DFS Pro datasets	Define fields, owners, refresh expectations, and dataset state.
Manage lifecycle	Dataset lifecycle	Control draft, review, published, retired, and replacement states.
Publish reporting output	BI reports	Confirm report owner, audience, refresh cadence, and access scope.
Monitor evidence	Audit and metrics	Track usage, changes, review activity, and data quality metrics.

The handoff should include dataset version, field dictionary, refresh cadence, permission scope, known exclusions, and report owner acceptance.

Handoff standard

Every workflow should end with a handoff record that a downstream team can use without reconstructing the path from memory. Include:

• source system and source owner;
• target site, asset, point, dataset, entity, or workflow identity;
• connector, mapping, sync, dataset, fusion task, or MDM decision reference;
• data freshness, units, time range, rejected-row status, and known quality limits;
• reviewer or steward decision;
• consuming workflow owner and accepted use;
• unresolved issues and the next review date when the data is still changing.

For DFS Lite-only workflows, the handoff can remain compact: connector, mapping, sync, quality note, and downstream owner. For DFS Pro workflows, include dataset version, lifecycle state, lineage, review queue status, and permission scope. For MDM workflows, include the master entity, aliases, steward decision, and any datasets that must be reprocessed after identity changes.

Related pages

Page	Use
Getting Started with DFS	Verify a small connector-to-quality loop before running a broader workflow.
DFS Lite	Connect, preview, map, sync, and check source data.
DFS Pro	Create governed datasets, fusion tasks, review queues, pipelines, and BI outputs.
DFS MDM	Govern master entities, aliases, steward queues, and identity corrections.
DFS Recipes	Follow task-oriented examples for common operating scenarios.
DFS Reference	Check connector types, mapping fields, permissions, and API surfaces.