ErgoTechWiki

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
====== MIStudio Architecture – Visual, Distributed, Intelligent ======

===== Overview =====

MIStudio takes the core idea behind TransSECS:

> If you can name it, you can control it

…and extends it into a **visual, distributed, real-time system for building industrial applications**.

But instead of writing code…

You connect blocks.

===== The Mental Model =====

Think of MIStudio as:

  * A **canvas of connected components**
  * Each component has:
    * **Inputs (Inports)** – what it receives
    * **Outputs (Outports)** – what it produces
  * You **wire them together** to create behavior

No hidden logic.

No black boxes.

Everything is visible and inspectable.

-----

===== MIStudio Overview =====

MIStudio is a drag-and-drop datops platform with visualization, vibe-coding, and connections that lets you build not just flows, but data products and views that can include AI/LLM blocks and easy-to-understand code blocks, including Python blocks with AI/ML libraries included.

It's a full data integration platform.

{{.:pasted:20260324-133237.png?1200| MIStudio Overview}}

<note>
Each block represents a capability:
  * MQTT publish/subscribe
  * REST calls
  * SPC calculations
  * AI/LLM interaction
  * Device communication
  * Custom logic
</note>

-----

===== How It Works =====

At runtime, MIStudio is:

  * A **network of components**
  * Passing **data in real time**
  * Reacting to events as they happen

Each connection represents:

  * A data flow
  * A trigger
  * Or both

-----

===== Data is the System =====

In MIStudio, there is no separation between:

  * Data
  * Logic
  * Integration

They are all part of the same flow.

Example:

  * Sensor data arrives via MQTT
  * It flows into an SPC block
  * Results trigger alerts
  * Alerts drive REST or PLC actions

All without writing traditional code.

-----

===== Works Across Protocols (Not Just One “Namespace”) =====

MIStudio is designed to work across **real industrial protocols**, including:

  * MQTT (pub/sub messaging)
  * AMQP (enterprise messaging)
  * Zenoh (edge/cloud distributed data)
  * OPC UA (industrial interoperability)
  * REST / HTTP APIs
  * SECS/GEM (semiconductor equipment)

Rather than forcing everything into a single abstraction, MIStudio:

> Connects them all together in a unified flow

-----

===== Example: Real-Time Data Flow =====

{{unsinlightsscreenshot.png?1200}}

What’s happening here:

  * Data is subscribed from MQTT topics
  * Aggregated and processed
  * Passed through analytics (SPC)
  * Published back out
  * Visualized or acted on

Each step is:

  * Visible
  * Configurable
  * Replaceable

-----

===== Built for Real Systems (Not Just Demos) =====

Many visual tools optimize for quick demos.

MIStudio is designed for:

  * **Production systems**
  * **Scalable architectures**
  * **Long-term maintainability**

Key differences:

  * Strong typing and structured data
  * Explicit connections
  * Reusable components
  * Integration with real protocols and equipment

-----

===== Integrated Intelligence (AI + Heuristics) =====

MIStudio treats AI as just another component in the flow.

{{.:pasted:20260324-132804.png?1200 | MIStudio AI Flow}}

You can:

  * Send data to an LLM
  * Interpret results
  * Feed them back into the system
  * Combine with rules and heuristics

Example use cases:

  * Extracting meaning from documents
  * Mapping data points automatically
  * Building dynamic control logic
  * Assisting operators

-----

===== Why This Architecture Is So Powerful =====

==== 1. Visual = Understandable ====

You can see:

  * What the system does
  * How data moves
  * Where decisions are made

No reverse engineering required.

-----

==== 2. Data + Logic + Integration in One Place ====

No need for:

  * Separate ETL tools
  * Separate automation engines
  * Separate integration layers

It’s all one system.

-----

==== 3. Works at the Edge and in the Cloud ====

Deploy anywhere:

  * Edge devices
  * On-prem systems
  * Cloud infrastructure
  * Containers / Kubernetes

-----

==== 4. Rapid Development Without Losing Control ====

You get:

  * Speed of low-code
  * Control of traditional engineering

-----

==== 5. Extensible by Design ====

New capabilities can be added as:

  * New components
  * New protocols
  * New integrations

Without rewriting existing systems.

-----

===== How to Think About MIStudio =====

A useful way to frame it:

> Node-RED is a great idea.

> MIStudio is what happens when you take that idea seriously for industrial, scalable, real-world systems.

-----

===== What This Means for You =====

With MIStudio, you can:

  * Build complete industrial applications visually
  * Integrate any data source or protocol
  * Add intelligence, rules, or AI anywhere
  * Adapt quickly as requirements change

All while maintaining:

  * Structure
  * Clarity
  * Control

-----

===== Summary =====

MIStudio is:

  * A **visual system builder**
  * A **real-time data orchestration platform**
  * An **integration layer across protocols**
  * A **foundation for intelligent automation**

Built on the same principle as TransSECS:

> If you can connect it, you can control it.

-----

===== Related Pages =====

  * [[mistudio:getting_started:start|Getting Started]]
  * [[mistudio:getting_started:what_is_mistudio|What is MIStudio?]]
  * [[mistudio:getting_started:project_structure|Project Structure]]
  * [[mistudio:getting_started:create_project|Create a Project]]
  * [[mistudio:getting_started:create_application|Create an Application]]
  * [[mistudio:getting_started:add_graphics|Add Graphics]]
  * [[mistudio:getting_started:edit_graphic_properties|Edit Graphic Properties]]
  * [[mistudio:getting_started:connect_data_source_to_graphic|Connecting a Data Source to a Graphic]]
  * [[mistudio:getting_started:test_data_sources|Using Test Data Sources]]
  * [[mistudio:design_editor:start|MIStudio Graphics]]