MaximaLabs
Live Digital Twin

Connect your plant, calibrate the model that already converged.

An edge gateway inside your OT network reads live tags read-only and ships them out over one outbound TLS connection. No inbound port, no write access to the DCS — just live data reconciled against the flowsheet you already solved.

Feed
Column
Distillate
Bottoms

Every converged stream and duty here is a live reconciliation target once tags are mapped — the twin compares this exact model against your plant's OPC-UA/Modbus feed.

One outbound connection, nothing inbound

The edge gateway reads OT tags read-only and buffers them on a network outage; the only connection leaving the OT/DMZ boundary is a single outbound TLS session authenticated with an API key — the same trust model IT/OT security teams already expect.

OTDMZCLOUDSensorsSCADAOPC-UA / Modbus / MQTTflowsim-edge-collectorReads tags, buffers on outageStore-and-forward bufferJSONL, survives outagesSingle outbound TLS + API keyLive digital twinSoft sensorsAnomaly detection

What the reconciled data buys you

Live readings compared against the solved model turn into inferred measurements and early warnings, not just a dashboard.

Deviation status per tag

Every mapped tag is classified ok, warn, alarm, or no-data by comparing the live reading against the simulated value — a clear signal, not a raw residual you have to interpret yourself.

Heat-exchanger fouling

Reconciling live duty and temperature data back to the exchanger's U-value surfaces fouling as it develops, ahead of the throughput hit.

Reconciled heat and mass balances

Live flow, temperature, and pressure readings are reconciled against the converged flowsheet so you can see where the plant is drifting from the model that was supposed to represent it.

PCA/Hotelling T² anomaly detection

A multivariate statistical model over the live tag stream flags abnormal operating states before they show up as an alarm.

Autoencoder anomaly detection

A second, independent anomaly detector — a hand-written autoencoder over the same tag stream — catches nonlinear deviations a linear statistical model would miss.

A reconciliation log you can audit

Every deviation, tag update, and reconnect is written to a streaming log, so a plant engineer can trace exactly when and where live data diverged from the model.

Fifteen minutes from install to live tags

The edge collector is a single gateway-VM service — no plant-wide integration project.

1

Install the edge collector

Deploy flowsim-edge-collector on a gateway VM inside your OT network — it speaks OPC-UA, Modbus, EtherNet/IP, and MQTT/Sparkplug B natively, no vendor driver required.

2

Map tags to the model

Point each live tag at the matching stream, component, or duty in your converged flowsheet — the same kind/edge/field mapping the simulator already uses internally.

3

Watch the twin reconcile

Live readings start flowing over the single outbound connection, the deviation matrix lights up in real time, and soft sensors and anomaly detectors start running against the stream.

Scripted from Python

Ingest readings and read the reconciliation from the flowsim SDK.

from flowsim.sdk import FlowSimClient

client = FlowSimClient()
client.add_twin_readings(sim_id, [
    {"tag": "COL.reboiler.duty", "value": 812_000.0},
])
status = client.twin(sim_id)  # live vs. simulated per-tag deviation

See it on your own process

Every number on this page comes from a real, converged flowsheet — open the workspace and run one yourself.

Stop fighting legacy software. Build your first flowsheet in 60 seconds.