Smart District Heating AI Tools

Smart District Heating uses product APIs for operator workflows and AI Engine endpoints for analytical computation. Available routes depend on the licensed module, deployment version, tenant configuration, and project integration.

Prerequisites

• The Smart District Heating module is enabled for the target tenant.
• Product API and AI Engine authentication are configured for the intended read or compute workflow.
• Network, station, asset, and source IDs have been validated.
• Input units, timestamps, quality flags, and model versions can be recorded.
• An operations or engineering owner is assigned to review output.

Tool layers

Layer	Primary use	Example namespace
Product API	Overview, stations, topology, history, anomalies, suggestions, operation logs, reports, and work records	/api/v1/heatops/*
AI Engine	State estimation, anomaly scoring, forecasts, network-twin computation, strategy preparation, and safety checks	/ai/heatops/*, /ai/mpc/*
DFS	Source connection, mapping, quality review, dataset preparation, and governed history	DFS APIs and workflows
Inspector or CMMS	Inspection, work-order execution, attachments, closure, and feedback	Project-configured work-order APIs

Use Smart District Heating in user-facing text. Use heatops only in paths, configuration keys, scopes, and permission strings.

Product API map

Workflow	Example routes	Review use
Overview and stations	/api/v1/heatops/overview, /api/v1/heatops/stations, /api/v1/heatops/stations/{stationId}, /api/v1/heatops/stations/{stationId}/history	Confirm station scope, status, freshness, and history.
Network	/api/v1/heatops/network/topology, /api/v1/heatops/network/leaks, /api/v1/heatops/network/hydraulic-analysis, /api/v1/heatops/network/pipeline/{pipelineId}/history	Review topology, suspected leaks, hydraulic context, and pipeline history.
Forecast	/api/v1/heatops/forecast, /api/v1/heatops/forecast/weather, /api/v1/heatops/ai/forecast/load, /api/v1/heatops/ai/forecast/supply-temp	Prepare load and supply-temperature review with weather context.
Anomaly and diagnosis	/api/v1/heatops/anomalies, /api/v1/heatops/anomalies/patterns, /api/v1/heatops/rca/faults, /api/v1/heatops/rca/diagnose/{stationId}	Triage findings and review diagnosis evidence.
Suggestions and logs	/api/v1/heatops/suggestions, /api/v1/heatops/suggestions/{suggestionId}/accept, /api/v1/heatops/suggestions/{suggestionId}/reject	Record operator review and connect accepted work to the operating log.

AI Engine map

Endpoint	Inputs	Output and review
POST /ai/heatops/state	Node readings, priors, zones, process and measurement settings	Estimated node and zone state, confidence, corrections, blockers, and evidence.
POST /ai/heatops/anomaly/detect	Station feature vector, outdoor temperature, cold-weather flag	Anomaly score, threshold, contributing features, and candidate fault type.
POST /ai/heatops/forecast/load	Historical load, weather, optional building coefficients, horizon	Load forecast, interval, model version, and available error context.
POST /ai/heatops/forecast/supply-temp	Historical supply temperature, weather, optional building coefficients, horizon	Supply-temperature outlook, interval, and model version.
POST /ai/heatops/twin/simulate	Mode, nodes, edges, stations, zones, observations	Four-layer hydraulic, thermal, station, and zone result with residual report.
POST /ai/mpc/strategy	Stations, current state, weather forecast, supply temperature, thermal inertia	Proposed supply target, ramp, valve adjustments, predicted temperatures, confidence, and safety report.
POST /ai/heatops/safety/check	Proposed action and approved constraint set	Passed, warnings, blocked values, checked constraints, and maximum severity.
POST /ai/heatops/safety/batch	Multiple actions and approved constraints	Per-action safety result for a reviewed decision package.

Recommended sequences

Daily operations

overview and station summary
  -> station and topology freshness
  -> load and supply-temperature forecast
  -> anomaly queue and diagnosis context
  -> reviewed suggestion
  -> operation log or work order
  -> measured outcome

Engineering scenario

versioned topology and input package
  -> /ai/heatops/state
  -> calibrated hydraulic and thermal study
  -> residual review
  -> alternative scenario
  -> safety and owner review
  -> monitored execution plan

Validation checklist

• Discover the available route set in the target environment.
• Confirm tenant, network, station, and asset IDs before every request.
• Record timestamp, unit, quality, input version, and model version.
• Treat state, anomaly, forecast, strategy, and simulation output as review evidence.
• Require project-approved constraints, authorization, audit, rollback, and manual override before write integration.

Failure handling

Problem	Response
Route is unavailable	Confirm module enablement, deployment version, tenant configuration, and integration contract.
State confidence is low	Review missing readings, priors, topology, and source quality before downstream use.
Forecast has insufficient history	Use a reviewed alternate method, show the limitation, and collect a longer representative period.
Simulation residuals fail acceptance	Revisit topology, units, observations, and calibration parameters.
Safety check blocks an action	Do not bypass it; revise the action or route it to the responsible engineering owner.