Facility Operations AI Tools
Facility Operations uses a shared AI tool surface to help agents and product workflows review assets, alarms, work orders, field inspections, energy records, facility analytics, and evidence packages.
Use this page when an Agent workflow or backend integration needs a code-backed map of the available tools and the expected review path.
Tool layers
| Layer | Primary users | Access boundary | Output type |
|---|---|---|---|
/mcp/base/ MCP tools | Agent clients, copilots, workflow orchestrators | base.read, base.compute.run, and selected base.action.write scopes | Asset context, documents, status summaries, data quality, analysis output, action drafts |
/api/v1/ai/* backend proxy | Facility Operations UI and backend services | Product service authentication and tenant permissions | Energy analysis, equipment knowledge, knowledge graph review, root-cause and impact summaries |
/ai/* AI Engine endpoints | Facility, SmartFM, energy, and FMS workflows | Product service authentication and workflow permissions | Energy profile, benchmark, Advisor chat, VAV diagnostic summary, chiller efficiency, work-order text extraction, simulation result |
/api/v1/* module APIs | Facility Operations UI and backend integration | Product RBAC scopes | Work orders, worker tasks, inspections, operations read model, asset center, model assets, energy, FMS, ESG, Green Mark evidence |
Use MCP tools when an Agent workflow needs governed access across operating records. Use AI Engine endpoints when the product workflow runs a controlled analysis, diagnosis, text-structuring, or simulation task.
MCP endpoint and scopes
| Endpoint | Scope | Use |
|---|---|---|
/mcp/base/ | base.read | Read asset context, equipment status, documents, connector state, compliance files, workflow tasks, and action-plan history. |
/mcp/base/ | base.compute.run | Run approved analysis such as energy, spatial, forecast, anomaly, simulation, optimization, and report generation tasks. |
/mcp/base/ | base.action.write | Create a work-order draft after the responsible operator approves the proposed action. |
Keep early Agent workflows read-only. Add write scope after the customer defines owner approval, audit capture, and rollback handling.
MCP tool map
Operating context and evidence
| Tool | Primary job | Typical question |
|---|---|---|
get_equipment_status | Read current equipment state, latest sensor readings, active alerts, and recent work orders. | "What is happening with this pump now?" |
get_equipment_documents | Retrieve manuals, SOPs, drawings, maintenance records, and related equipment documents. | "Which SOP should the technician check?" |
search_documents | Search ECM documents by keyword, type, or related entity. | "Find the latest inspection procedure for this room." |
query_knowledge | Query equipment types, failure modes, repair actions, diagnostic rules, and maintenance schedules. | "What are common causes for this alarm?" |
get_action_plan_history | Review previous AI action plans, workflow approval status, execution outcomes, and linked documents. | "Have we tried this fix before?" |
get_pending_tasks | Read approval, review, or signature tasks from the workflow system. | "Which maintenance reports are still awaiting approval?" |
get_compliance_documents | Retrieve certificates, audit reports, and evidence packs by standard or topic. | "Show evidence documents related to this clause." |
get_expiring_documents | Find documents approaching retention end date or periodic review. | "Which facility documents need review this month?" |
Data and connector readiness
| Tool | Primary job | Typical question |
|---|---|---|
list_connectors | List configured data connectors with status, source, and last sync time. | "Which BMS or CMMS connectors are active?" |
check_data_quality | Review completeness, accuracy, consistency, timeliness, and top violations. | "Is this dataset ready for an operations summary?" |
troubleshoot_connector | Diagnose connector errors and recent sync logs. | "Why are meter readings stale?" |
import_data | Import and process external REST or CSV records for analysis. | "Load this exported maintenance history for review." |
Facility analysis
| Tool | Primary job | Typical question |
|---|---|---|
calculate_emissions | Calculate greenhouse-gas emissions by scope with regional default factors. | "What is the estimated carbon footprint for this period?" |
analyze_spare_parts | Review spare-parts usage, slow-moving inventory, stockout risk, and reorder guidance. | "Which parts need procurement attention?" |
forecast_timeseries | Forecast time-series values using approved forecasting methods. | "How may this load trend change next week?" |
detect_anomaly | Score data points for anomalies. | "Which readings look unusual?" |
fit_distribution | Fit probability distributions to observed failure, service, or arrival data. | "What distribution describes these repair durations?" |
find_changepoints | Detect structural breaks in a time series. | "When did this energy pattern change?" |
generate_report | Generate a status, alert, equipment, or simulation report. | "Create a reviewed summary for this shift handover." |
Spatial and facility movement
| Tool | Primary job | Typical question |
|---|---|---|
analyze_spatial_anomaly | Detect unusual values in spatial heatmap sensor data. | "Where are the hot spots on this floor?" |
compare_zones | Compare statistics between two zones or floors. | "How does floor 1 compare with floor 2?" |
recommend_sensor_placement | Recommend sensor locations based on coverage gaps and confidence. | "Where should we add sensors?" |
find_path | Find navigation paths across a building with multi-floor and accessibility options. | "How should a worker reach this equipment?" |
Action drafting
| Tool | Primary job | Typical question |
|---|---|---|
create_work_order | Create a work-order draft from an approved recommendation. | "Draft the work order for the technician." |
create_work_order should be reached only after the workflow has the correct write scope and an operator has approved the action text, owner, location, asset, priority, and evidence references.
AI Engine and backend proxy tools
| Endpoint | Purpose | Inputs to prepare | Output to review |
|---|---|---|---|
/api/v1/ai/energy/analyze and /ai/energy/analyze | Analyze facility or equipment energy readings. | Meter mapping, time window, asset context, operating schedule, weather or load context when available. | Anomaly notes, recommendations, carbon estimate, and evidence references. |
/api/v1/ai/energy/profile and /ai/energy/profile | Build an equipment energy profile. | Equipment ID, recent readings, baseline window, equipment type, operating mode. | Profile summary, typical range, anomaly indicators, and recommended checks. |
/api/v1/ai/energy/benchmarks and /ai/energy/benchmarks | Compare energy performance against benchmark references. | Building or facility type, region, time window, normalized consumption inputs. | Benchmark category, variance, and review notes. |
/ai/advisor/chat | Run a facility management assistant conversation with optional facility context. | User question, tenant context, asset or alert context, evidence snippets. | Answer draft, tool-use trace, missing evidence, and suggested next checks. |
/api/v1/ai/knowledge/equipment/{type} | Retrieve equipment knowledge by type. | Equipment type and language preference. | Failure modes, maintenance procedures, and knowledge snippets. |
/api/v1/ai/graph/root-cause | Review knowledge graph context for possible root causes. | Alert, asset, related systems, inspection findings, work records. | Root-cause candidates and supporting graph paths. |
/api/v1/ai/graph/impact | Review impacted systems or assets. | Asset or event context, graph relationships, operating boundary. | Impacted assets, systems, and review notes. |
/ai/llm-pdm/structure-workorder | Structure free-text maintenance or inspection feedback. | Raw text, equipment context, work-order template, language. | Structured fields, extracted findings, confidence, and reviewer corrections. |
/ai/hvacops/vav/diagnostics/summary | Summarize VAV diagnostic state. | VAV readings, setpoints, airflow, temperature, time window. | Diagnostic summary, affected zones, and recommended engineering checks. |
/ai/hvacops/chiller/efficiency | Review chiller efficiency. | Chiller load, COP inputs, ambient condition, operating state, equipment metadata. | Efficiency summary, comparison notes, and operating review input. |
/ai/sim/extract-kpis, /ai/sim/build-idf, /ai/sim/run | Prepare and run EnergyPlus-oriented building simulation workflows. | Building geometry, zones, materials, schedules, weather, meter data, operating assumptions. | KPI list, IDF payload, run status, simulation outputs, and assumptions. |
Recommended tool sequences
Daily operations triage
list_connectors
-> check_data_quality
-> get_equipment_status for affected assets
-> search_documents or get_equipment_documents for supporting procedures
-> generate_report for shift review
Use this sequence when the operator needs a practical summary of current issues, source freshness, and next checks.
Alert-to-work-order review
get_equipment_status
-> query_knowledge for failure modes and checks
-> get_action_plan_history for prior decisions
-> /api/v1/ai/graph/root-cause or Advisor chat
-> create_work_order after owner approval
The reviewed output should separate confirmed evidence, likely checks, missing data, and the draft work-order fields.
Energy and sustainability review
list_connectors
-> check_data_quality
-> /ai/energy/profile or /ai/energy/analyze
-> calculate_emissions when carbon input is requested
-> generate_report for internal review
Use this sequence for meter-driven review, ESG preparation, and facility performance discussions. Confirm tariff, weather, meter mapping, and operating schedule assumptions before external use.
Green Mark evidence preparation
get_compliance_documents
-> get_expiring_documents
-> search_documents for clause evidence
-> check_data_quality for source records
-> generate_report or evidence-pack assembly in the product workflow
Use this sequence to organize source evidence, reviewer notes, and missing items for project owners.
FMS diagnostic review
get_equipment_status
-> get_equipment_documents
-> query_knowledge
-> /ai/hvacops/vav/diagnostics/summary or /ai/hvacops/chiller/efficiency
-> reviewed maintenance or engineering summary
Use this sequence for chiller, HVAC, VAV, and subsystem reviews when site data is mapped and current.
Data requirements
Facility Operations AI tools are strongest when the data package includes:
- stable tenant, site, floor, zone, space, and equipment IDs;
- source-system IDs for BMS, CMMS, inspection, meter, ESG, and evidence systems;
- timestamped sensor readings, meter readings, alerts, inspection records, and work-order updates;
- source freshness and DFS quality notes;
- equipment manuals, SOPs, drawings, maintenance records, and compliance evidence;
- meter units, carbon factors, tariff assumptions, weather context, and operating schedule where relevant;
- model asset and twin-model binding IDs when spatial context is required;
- approval owner for work-order drafts, evidence packs, and externally shared reports.
Use Connect BMS Data to a Facility Twin and Fuse Inspection, Work Order, and Sensor Data before workflows that depend on fused operational evidence.
Output review checklist
- Tenant, site, asset, meter, and source-system IDs are visible.
- Connector state and data freshness are included when source data is used.
- Work-order and inspection records include owner, status, timestamp, and audit reference.
- Energy and carbon outputs show assumptions, time window, units, and source mapping.
- Diagnostic output is linked to supporting readings, documents, and operating context.
- Green Mark or ESG evidence remains in an internal review path until project owners approve external use.
- Action drafts remain separate from approved work orders.
- Operator corrections and work-order outcomes are captured for later workflow improvement.