ZeroPipeline.Nodes 1.0.0

ZeroPipeline

ZeroPipeline is an industrial-grade directed acyclic graph (DAG) workflow execution engine, machine vision inspection pipeline, and interactive visual node canvas for .NET with zero external dependencies. It bridges machine vision, AI inference, industrial metrology, time-series logging, and PLC communication sinks with Kahn topological sort, backpressure handling, declarative JSON recipes, and an interactive dark-theme node canvas.

🌟 Architecture & Sub-Packages

graph TD
    UI["ZeroPipeline.UI (Visual Canvas & Studio)"]
    Recipe["ZeroPipeline.Recipe (JSON Schema & Node Registry)"]
    Nodes["ZeroPipeline.Nodes (Vision, Metrology, TSDB, PLC)"]
    Core["ZeroPipeline.Core (DAG, Kahn Sort, Backpressure Ports)"]

    UI --> Recipe
    UI --> Nodes
    Recipe --> Core
    Nodes --> Core

📦 Installation

Install via the .NET CLI:

dotnet add package ZeroPipeline.Core
dotnet add package ZeroPipeline.Nodes
dotnet add package ZeroPipeline.Recipe
dotnet add package ZeroPipeline.UI

🚀 Quick Start

1. Building and Running a Pipeline DAG in Code

using ZeroPipeline.Core.Graph;
using ZeroPipeline.Core.Execution;
using ZeroPipeline.Nodes.Vision;
using ZeroPipeline.Nodes.Inspection;
using ZeroPipeline.Nodes.Storage;

// 1. Construct the pipeline graph
var graph = new PipelineGraph();

var camera = new SyntheticCameraNode("Cam01");
var caliper = new MetrologyCaliperNode("Caliper01", expectedWidthMm: 30.0, toleranceMm: 0.1);
var tsdb = new TsdbStorageNode("Tsdb01");

graph.AddNode(camera);
graph.AddNode(caliper);
graph.AddNode(tsdb);

// Wire ports (Camera Frame -> Caliper Image, Caliper Result -> TSDB Sink)
graph.Connect(camera.Outputs[0], caliper.Inputs[0]);
graph.Connect(caliper.Outputs[0], tsdb.Inputs[0]);

// 2. Instantiate and run pipeline executor
var executor = new PipelineExecutor(graph);
await executor.InitializeAsync();

// Execute a single discrete cycle
await executor.ExecuteStepAsync();

2. Declarative JSON Recipe Persistence

using ZeroPipeline.Recipe.Builder;
using ZeroPipeline.Recipe.Serialization;

// Export active graph to declarative JSON
var builder = new RecipeGraphBuilder();
var recipe = builder.ExportRecipe(graph, "recipe_aoi_01", "AOI Metrology Routine");
string json = RecipeJsonSerializer.Serialize(recipe);

// Reconstruct pipeline dynamically from JSON without recompilation
var loadedRecipe = RecipeJsonSerializer.Deserialize(json);
var reconstructedGraph = builder.BuildGraph(loadedRecipe);

3. Embedding the Visual Pipeline Studio in WinForms

using ZeroPipeline.UI.Studio;

var studio = new ZeroPipelineStudioControl
{
    Dock = DockStyle.Fill
};

// Bind to live execution engine
studio.AttachExecutor(executor);

// Load recipe into visual canvas
studio.LoadRecipeJson(json);
this.Controls.Add(studio);

📊 Benchmark & Performance

Tested on Intel Core i7-13700K (.NET 8.0, Release x64):

📄 License

MIT License © 2026 Phong Võ. Part of the ZeroPlatform project.