← Digital Twin Engine

Digital Twin Engine

Model Composition

Individual equipment models compose into process-level directed acyclic graphs (DAGs). Wire typed ports together, let the engine resolve topological execution order, and propagate state through the entire process graph every 100ms.

EQUIPMENT MODELSPump Modelη = P_hyd/P_e3in2outBlower ModelP = Q·Δp/η2in2outClarifierSOR = Q/A3in3outFilterΔP = μ·R·Q/A2in2outCOMPOSITION CANVASflow, headairflow, DOTSS, BODeffluent1Pump Modelη = P_hyd/P_e2Blower ModelP = Q·Δp/η3ClarifierSOR = Q/A4FilterΔP = μ·R·Q/Atopo_sort:[1] pump[2] blower[3] clarifier[4] filterPROCESS OUTPUTSTATE VECTORSystem Efficiency84.2 %Total Flow142.5 L/sEnergy Cost0.38 $/m³Process Capacity91.3 %Effluent Quality73.1 %composition-engine@twinedge:~$compose> add pump_model pipe_model clarifier filterDAG valid | No cycles | 4 models | 6 edges | Execution order resolvedEXEC TIMELINE:t0: pump t1: blower t2: clarifier t3: filter tick: 0MODEL COMPOSITION DAG -- TWINEDGE DIGITAL TWIN ENGINE

How Composition Works

Three stages from isolated models to a unified process simulation.

01

Define Models

Select from the built-in library or create custom equations. Each model declares typed input and output ports with explicit units and data types.

02

Wire Connections

Connect output ports to input ports on the composition canvas. The engine validates type compatibility between ports and detects cycles in the graph.

03

Execute the DAG

The engine resolves topological order from the graph structure and propagates state through every node every 100ms. Outputs cascade downstream automatically.

Composition Capabilities

The graph execution engine that powers every composed process model.

Automatic Dependency Resolution

The engine determines execution order from the graph topology. Add or remove models and the topological sort recomputes instantly — no manual sequencing.

Parallel Branch Execution

Independent sub-graphs execute concurrently across available cores. Branches with no data dependencies run in parallel for minimum propagation latency.

Type-Safe Connections

Port types — flow, pressure, temperature, concentration — are checked at composition time. The engine rejects incompatible connections before the graph starts.

Hot-Swap Models

Replace a model in the running graph without stopping the simulation. The engine re-resolves dependencies and continues propagation from the next tick.

Process Templates

Pre-built DAG compositions for common industrial process trains.

TemplateModelsConnectionsIndustryLatency
Wastewater Treatment Train812Water/WW<80ms
HVAC Chiller Plant57Commercial<50ms
Production Line69Manufacturing<60ms
Pump Station34Water/WW<30ms

Compose Your First Process

Wire equipment models into a process-level DAG. The engine resolves execution order, validates port types, and propagates state through the graph in under 100ms.