Introduction

MACHHUB is a self-hostable Industrial IoT (IIoT) platform and unified data fabric. It gives manufacturing and operations teams one place to collect, model, store, automate, and serve their data — and gives developers a type-safe SDK to build applications on top of it.

The whole product ships as MACHHUB Platform: a single self-contained service that runs at any level of your deployment stack — from a Raspberry Pi on the plant floor all the way up to an enterprise server or a cloud VM — and bundles everything you need.

Deployment level	Typical host
Edge	Raspberry Pi, industrial PC, PLC gateway
On-line	Line-side server, MES host
On-site	Site/plant-level server, on-premise rack
Enterprise	Corporate data center
Cloud	Cloud VM or managed container

Who is it for?

• Administrators manage users and domains through the web console.
• Application developers build data models (Collections) using the console, then use the @machhub-dev/sdk-ts SDK and the MACHHUB Designer VS Code extension to build dashboards, line-side apps, and Manufacturing Execution System (MES) applications.
• Integrators & device engineers wire the Unified Namespace, enable Historian on tags, publish and subscribe to live Tags over MQTT, and bridge to other MACHHUB instances with Upstreams.

The building blocks

Concept	What it is
Domain	A tenant/workspace. Data, users, groups, and namespaces all belong to a domain. The built-in admin domain is domains:machhub_admin.
Application	A user-developed app you build and deploy on MACHHUB, with its own runtime (start/stop, served on a port). Today each Application is backed by an application-type domain; the two are being separated.
Collection	A typed data table (schema + records) with fields, relations, and indexes. Think “your database, modeled in the console.”
Unified Namespace (UNS)	A hierarchical tree of namespaces → folders → tags that mirrors your plant. Every tag maps to an MQTT topic.
Tag	A single live signal (a leaf in the UNS). You publish/subscribe to its value over MQTT in real time.
Historian	Time-series storage. Any tag can be historized (on-change or sampled) with a retention policy, then queried or exported.
Process	A serverless function (Python or TypeScript) that runs on the platform. Triggered by a schedule, a tag change, an HTTP call, or manually.
Flow	A Node-RED visual pipeline for ingesting data from physical devices — PLCs, sensors, and other sources not natively supported by MACHHUB. Flows use the @machhub-dev/node-red-nodes library to read/write Tags and interact with Collections and the database, bridging device data into the platform.
Upstream	An MQTT bridge that mirrors parts of your UNS to (or from) another MACHHUB instance (e.g. an edge MACHHUB up to a central one).
SDK	The official TypeScript client, @machhub-dev/sdk-ts, for building apps against all of the above.
Designer	A VS Code extension that connects to a MACHHUB Platform (your MACHHUB Environment), manages source/builds in a sidebar, deploys your project from the editor, and proxies your dev server’s SDK requests to the platform.

Processes vs. Flows

These are two different features. Processes are serverless functions you write in code; Flows are Node-RED visual pipelines. See Processes & Flows for the full comparison.

How the pieces fit together

flowchart TB
  subgraph EDGE["MACHHUB Platform"]
    API["REST API"]
    MQTT["MQTT Broker (UNS / Tags)"]
    DB[("Data store + Historian")]
    Procs["Processes (Python / TS runtimes)"]
    UIsTATIC["Serves the Web Console"]
  end

  Devices["PLCs / sensors / gateways"] -- "native MQTT" --> MQTT
  Devices -- "OPC-UA / Modbus / serial / etc." --> Flows["Node-RED Flows\n(@machhub-dev/node-red-nodes)"]
  Flows -- "ingest → write tags / collections" --> API
  Console["Web Console (browser)"] -- "REST + MQTT-over-WS" --> API
  App["Your app (SDK)"] -- "REST + MQTT" --> API
  Designer["VS Code: MACHHUB Designer"] -- "proxy dev SDK + deploy project" --> API

  API <--> DB
  MQTT --> DB

What you can build

• Real-time dashboards that subscribe to tags and render live values and charts.
• OEE / DAQ / quality apps backed by Collections and the Historian.
• Manufacturing Execution System (MES) apps — work orders, traceability, and production tracking on Collections, tags, and the Historian.
• Warehouse Management System (WMS) apps — inventory, locations, and receiving/picking workflows backed by Collections.
• Automations — alerting, data transforms, ERP/MES integrations — as Processes.
• Device ingestion pipelines — bridge PLCs, sensors, and unsupported protocols into the platform via Node-RED Flows.
• Edge-to-cloud pipelines that bridge selected tags upstream over MQTT.

Where to go next

• New to the platform? Continue with the Quickstart.
• Want the mental model first? Read Architecture.
• Self-hosting? Jump to Install & Self-Hosting.
• Building an app? Go to SDK initialization.