Platform

The decision layer your plant systems were missing.

ERP plans. MES schedules. SCADA measures. None of them decides. SmeltIQ sits between enterprise planning and shop-floor execution, stitches their data into one live operating view, and runs a governed loop on every heat: check, recommend, explain, learn.

The industrial stack

Three layers, one operating view.

All three copilots run on the same platform. Add a copilot and it inherits every connection, every rule and every heat of history already in the system.

Enterprise & planning

ERP / SAPOrders, grade plan, material master
MES / SchedulingHeat plan, route, sequence, delays
LIMS / LabChemistry, samples, release status
SmeltIQ decision layerRecipe Copilot · Furnace Copilot · Data Copilot
Checks
Recommends
?Explains
Learns
SCADA / PLCTemperature, power, oxygen signals
InventoryAdditive lots, price, availability
QMS / TraceabilityDeviations, approvals, audit trail

Shop floor & material

Your systems tell you what happened.

SmeltIQ helps decide what happens next.

What connects, and what it brings.

The recommendation is only as good as the context behind it. Each system adds a piece of the picture that a spreadsheet never had.

ERP / SAPOrders, grade plan, cost centres, material master. PM / PP / CO modules for batch traceability, maintenance logging and cost variance.
MES / SchedulingHeat plan, route, sequence and delay codes. Native connectors to multiple MES modules.
LIMS / LabChemistry, sample history and release status — the moment the lab posts it.
SCADA / PLC / DCSTemperature, power, oxygen and runtime signals in; recommendations and alerts back out to the console.
InventoryAdditive availability, lot, price and supplier — so the recommendation is buildable and costed.
QMS / TraceabilityDeviations, approvals and the full audit trail of every decision.
Cameras and acousticsIndustrial camera and microphone feeds from the furnace, in extreme environments.
ESG frameworksAutomated emissions telemetry to SASB, CDP, UN SDGs, ISO and SBTi reporting.

Governance is a feature, not a policy document.

A recommendation the operator cannot question will be ignored. A system leadership cannot audit will not be trusted. The platform is built so both can happen on every heat.

Approvals and reason codes

Every recommendation is approved, adjusted or overridden by a person, and every deviation carries a reason.

Audit trail

Inputs, model version, confidence, the action taken and the outcome — linked, for every heat, forever.

Model confidence

The copilot says how sure it is and on how many similar heats. Low confidence falls back to your rules.

Fallback rules and change control

Your existing formulas stay as the safety net. Model changes are versioned, tested and released like any other plant change.

Deployed the way industrial plants need.

Cloud platform for models and learning; direct, secure integration for the data that must stay close to the process.

Cloud platform

Models, training loops and dashboards on SmeltIQ's cloud platform, with direct integration to plant systems.

Secure API and upload

Camera images and process data arrive by direct integration or secure API-based upload — no manual handling.

Console, mobile HMI, SCADA/DCS

Insights land where the shift already looks: operator consoles, mobile HMI devices, or straight into SCADA and DCS.

Scales across furnaces

Adapts to different furnace types and industrial needs without additional workforce — add a furnace, not a team.

Questions we get asked

Platform and integration, answered.

What IT, OT and automation teams ask before connecting SmeltIQ to plant systems.

What is the SmeltIQ decision layer?

It is the layer between enterprise planning and shop-floor execution. ERP, MES and LIMS feed it business, production and chemistry context; SCADA, PLC, inventory and QMS feed it live process, material and traceability data. The layer checks, recommends, explains and learns on every heat, and all three copilots run on it.

Which systems does SmeltIQ connect to?

ERP and SAP (PM, PP, CO), MES and scheduling systems, LIMS and lab systems, SCADA, PLC and DCS, inventory systems, QMS and traceability systems, maintenance and CMMS systems, and industrial camera and acoustic feeds.

How is AI governance handled in a regulated plant?

Through approvals on every recommendation, reason codes on every deviation, a complete audit trail linking inputs, model version, confidence, action and outcome, explicit fallback rules when confidence is low, and versioned change control for model releases.

Is deployment cloud, on-premise or hybrid?

Models, training loops and dashboards run on the SmeltIQ cloud platform with direct secure integration to plant systems, and insights are delivered to operator consoles, mobile HMI or straight into SCADA and DCS. Data capture at the furnace runs at the edge.

How does it scale across multiple furnaces or plants?

Adding a furnace or a copilot inherits the existing connections, rules and accumulated heat history rather than starting a new project, and does not require additional workforce.

Show us your systems map.

Two weeks of data and recipe audit is enough to tell you what connects, what is missing, and where the first copilot should go.