OIPF / generic-industrial-process

Generic Industrial Process Library is the first ProcessGit library for OIPF — the Operational Intelligence Packaging Format.

It is not a normal software repository and not a general documentation site. It is a ProcessGit process library intended to be published as:

https://processgit.org/OIPF/generic-industrial-process

The library provides a reusable, industry-neutral production process skeleton for AI-native operational intelligence projects. It defines the common process pattern by which an industrial site is observed, contextualised, modelled, recommended, acted upon, recorded, and improved.

Purpose

Most industrial AI deployments become bespoke because every plant has different tags, historians, maintenance records, recipes, operator workflows, and local terminology.

This ProcessGit library provides a neutral starting point for repeatable industrial intelligence packages:

Observe → Contextualise → Infer → Check → Recommend → Approve → Act → Record → Learn

It is designed to support future OIPF packages such as:

• cleaning optimisation;
• fouling detection;
• energy-loss detection;
• predictive maintenance;
• quality drift detection;
• downtime root-cause analysis.

What this repository contains

.processgit/                 ProcessGit metadata hints
processgit.yaml              ProcessGit library descriptor
processgit.viewer.json       Viewer/editor hints for ProcessGit
library.yaml                 OIPF process library manifest
levels/                      UAPF-style L0-L4 process decomposition
bpmn/                        BPMN model for the generic industrial process
/dmn                         Decision model example
process-cards/               Process cards for the library
algorithm-cards/             Algorithm cards referenced by process tasks
packages/                    First OIPF package skeletons
resources/schemas/           OIPF schemas used by packages
resources/examples/          Minimal deployable examples
docs/                        Concept, architecture and publishing notes

ProcessGit classification

Recommended ProcessGit repository metadata:

repo_type: process
uapf_level: 0-4
status: draft
reference_kind: n/a

If your ProcessGit instance stores classification at platform level, set these values in the repository settings rather than treating this file as the source of truth.

Core idea

The library treats production operations as versioned process assets:

• assets and lines;
• industrial tags;
• operating states;
• events;
• constraints;
• model contracts;
• recommendations;
• human approvals;
• outcomes;
• feedback loops.

It does not replace OPC UA, ISA-88, ISA-95, MQTT/Sparkplug, MES, CMMS, ERP, SCADA, DCS, or historians. It defines an AI-native process layer above them.

First package

The first concrete package is:

packages/cleaning-optimisation/

It is intentionally simple. It is meant to become the first real OIPF package that can be evolved into a deployable production-intelligence process.

Publishing

To publish to ProcessGit:

git clone https://processgit.org/OIPF/generic-industrial-process.git
cp -R generic-industrial-process/* generic-industrial-process/.processgit generic-industrial-process/.github <repo>/
cd <repo>
git add .
git commit -m "Initial OIPF generic industrial process library"
git push origin main

To mirror to GitHub, create a repository such as:

https://github.com/OIPFormat/generic-industrial-process

and push the same contents.

Status

Draft reference library. Not yet a normative OIPF specification.