ARG INDUSTRIAL DEPLOYMENT

From plant datato Digital Twin & AI

Three complementary documents for refineries, petrochemical plants and factories: assess the existing site, connect industrial data and deploy AI infrastructure. Start with real site information and carry the plan through acceptance testing and operations handover.

Explore the three guides Download centre

1Site assessment & implementationPrintable assessment form
2Connect the software to plant systemsOT integration reference architecture
3Deploy AI for the digital twinARG_AI deployment engineering basis

OT → DIGITAL TWIN → AI

Data, analysis and decisions, with clear boundaries

This path connects each observation to its source and quality, asset identity, model version and answer references.

01Plant sourceDCS / PLC · Historian
02Approved publicationCollector · OT / DMZ
03Twin stateAsset / ENID · time & quality
04Analysis & knowledgeML · RAG · approved sources
05Evidence-led explanationLLM · agent · evidence citations
06Operator decisionHuman review · site procedure

In these documents, data arrives through the approved OT/DMZ path, process control remains in the DCS/PLC, and AI results support review and decisions by the engineer or operator.

PROJECT DOCUMENTS

A clear role for every document

Complete the assessment with the site team, then use the companion engineering references to design connectivity and AI infrastructure.

Printable assessment form1

Site assessment & implementation

Turn real site information into a defined package for project design and delivery.

PDF · 13 pages

Who is it for?

Client engineering, IT/OT, process engineers and the implementation team

Record use cases, equipment, data sources and Tag-to-Asset-to-ENID mapping
Survey networks, access, infrastructure capacity and AI prerequisites
Prepare the ICD, equipment list, schedule, responsibilities and FAT/SAT

Expected outcome

A site baseline for detailed engineering and acceptance planning, completed by the project teams.

OT integration reference architecture2

Connect the software to plant systems

Design the data path from the existing control system to the asset model in the digital twin.

Digital Twin integration with refinery & petrochemical control systems

PDF · 32 pages

Who is it for?

Automation, instrumentation, network and OT security engineers

Identify DCS/PLC, the historian and data publication interfaces such as OPC UA
Define the collector, approved OT/DMZ path and value, time and quality mapping
Plan installation, load tests, outage recovery and integration acceptance

Expected outcome

A testable path for approved telemetry reads; process control stays in the existing DCS/PLC.

ARG_AI deployment engineering basis3

Deploy AI for the digital twin

Connect validated equipment data and organisational knowledge to predictions, cited retrieval and traceable analysis.

PDF · 28 pages

Who is it for?

Infrastructure, data science, software and operations teams

The ML lifecycle: data preparation, training, inference, evaluation and model monitoring
Build RAG from approved documents, enforce access and return cited answers
Deploy a local LLM or connect provider APIs and orchestrate agent tools

Expected outcome

An AI deployment and validation plan; final sizing is benchmarked against the selected model, site data and user load.

Data, analysis and decisions, with clear boundaries

A clear role for every document

Site assessment & implementation

Connect the software to plant systems

Deploy AI for the digital twin

Explore the architecture. Prepare the project.