GFGFIS Evidence PortalChatake Innoworks · Version 3
Next build · software-in-the-loop plant

A real 3D anaerobic-digestion plant driven by the same GFIS process state.

The 3D plant will not be decorative animation. Tank level, gas holder, pipes, equipment, alarms and sensors will respond to one synchronized ADM1/GFIS state and a versioned telemetry stream.

Yes, this can be built and connected

What is technically achievable now

A browser-based 3D software-in-the-loop plant can receive virtual IoT telemetry from GFIS, show equipment state, animate gas/liquid behaviour, trigger alarms and preserve every simulation event.

What requires a physical plant later

A synchronized industrial digital shadow requires sensors, PLC/SCADA access, calibration, secure networking and safety approval. GFIS should initially remain read-only and advisory.

Operator or replay scenario
Safety and constraint engine
ADM1 + GFIS hybrid state
Virtual IoT and historian
3D plant + Control Room

Every visible component has a state variable

Feedstock receptionMass, moisture, TS, VS, contamination and delivery queue
PreprocessingSorting, shredding, dilution, pump state and energy demand
Digester tankLevel, temperature, pH, OLR, HRT, mixing and stability colour
Gas holderVolume, pressure, CH₄/CO₂/H₂S and inflation state
Gas treatmentDesulfurisation, drying, methane slip and media life
Biomethane upgradingFlow, recovery, purity, energy and off-spec diversion
Compression/storageCompressor state, cascade pressure, temperature and ESD state
CHP or injectionElectricity, heat, gas dispatch, metering and export route
Digestate systemStorage, nutrient output, treatment and dispatch
Flare and reliefEmergency condition, gas diversion and emissions record
Valves and pumpsOpen/closed state, flow, current, fault and maintenance timer
Sensor layerValue, unit, quality, timestamp, calibration and communication state

Virtual IoT telemetry contract

Signal familyExample signals3D responseGFIS interpretation
Biological processTemperature, pH, feed, level, OLR, HRTTank colour, liquid level, agitator and feed animationMethane prediction and stability state
Gas systemFlow, pressure, CH₄, CO₂, H₂S, moistureGas-holder inflation, pipe flow, quality and alarm colourYield, quality, treatment and dispatch decision
EquipmentMixer RPM/current, pump state, valve positionMotion, state badge and fault animationEnergy use, anomaly and maintenance evidence
SafetyLEL, high pressure, ESD, flare stateRed alarm, isolation path and emergency timelineSafety event only; local PLC/ESD remains authoritative
AI/twinPredicted methane, VFA/ALK, confidence, physics flagForecast overlays and explanatory panelsAdvisory decision and evidence memory

Three stages of realism

Stage 1

Software-in-the-loop virtual plant

ADM1/GFIS generates virtual telemetry and drives the 3D scene. This is the defensible next M.Tech/product milestone.

Next build
Stage 2

Read-only digital shadow

A physical plant streams continuous IoT data into the historian and 3D plant. GFIS recommends; it does not control safety equipment.

Requires plant access
Stage 3

Operational digital twin

Calibrated live synchronization, governed recommendations and carefully approved control integration with PLC/ESD authority.

Long-term validation

Software architecture

Process engineADM1 or reduced-order mass-balance state plus GFIS methane/stability models.
Telemetry brokerVersioned virtual sensor messages; later MQTT/OPC UA/SensorThings adapters.
Twin state storeTimestamp, plant ID, units, quality, alarm, model version and scenario lineage.
3D rendererBrowser WebGL scene with selectable equipment, overlays and time controls.
Control RoomTrends, explanations, scenarios, export, replay and operator notes.
Safety boundary: emergency shutdown, gas compression protection and plant interlocks remain in certified local PLC/ESD systems. A research/cloud twin must never replace them.

First buildable milestone

1

One digester

Tank, feed, mixer, gas holder, treatment, compressor, storage and digestate system.

2

One shared state

GFIS process engine publishes virtual telemetry with explicit units and evidence labels.

3

Five scenarios

Nominal, overload, pH shock, low temperature and corrective recovery.

4

Visible causality

Each slider change shows which variables, equations, alarms and outputs changed.

5

Replayable memory

Every run exports telemetry, events, predictions, physics checks and screenshots.

6

Validation boundary

Label it a virtual plant until physical sensor synchronization is demonstrated.