IronFlock Documentation

Industrial AI and digital operations should not require months of integration work. IronFlock is a middleware system that provides the missing layer between your machines and the software that makes them intelligent — connectivity, real-time data, AI execution, secure access, and app delivery, all in one architecture.

The Problem IronFlock Solves

Every industrial team building digital services for machines runs into the same challenge: before you can deploy a monitoring dashboard, an AI model, or a remote service tool, you first need to build an integration stack — device connectivity, data pipelines, authentication, secure tunneling, container orchestration, and more. That middleware work takes months and pulls engineering resources away from the actual application.

IronFlock eliminates that layer. You connect your machines, PLCs, sensors, and gateways, then focus entirely on the applications and AI workflows that create value.

What IronFlock Is

Think of IronFlock like Android on a smartphone — it stays behind the scenes while users interact with your apps and your brand. The complexity lives in the middleware, not in the day-to-day experience.

At its core, IronFlock is an industrial middleware system that connects multiple capabilities into one runtime:

• Device & Machine Connectivity — Register machines, sensors, PLCs, and gateways. Organize them into projects, monitor connectivity, and push over-the-air updates across sites.
• App Development & Deployment — Write code in any language, package it with Docker, and deploy to edge devices or cloud. Use the built-in cloud IDE or your own tools with GitHub/GitLab integration.
• Real-Time Data Collection — Automatically provision per-project time-series databases. Collect telemetry from devices and visualize it through dashboards — with sub-second latency from edge to screen.
• Custom Boards — Build real-time visualizations, control panels, and multi-page dashboards using a no-code widget system. Charts, gauges, maps, tables, forms, and device controls — all defined in a simple YAML template and updated live as data arrives.
• Multi-Agent AI Orchestration — Each app can ship its own specialized AI agents for different tasks — anomaly detection, root cause analysis, process optimization, natural language queries over machine data. IronFlock orchestrates them through one runtime with shared conversations, permissions, and operational context.
• Secure Remote Access — Tunnel into device web interfaces, HMIs, camera streams, PLC diagnostics, or remote desktops without opening ports. Every session is logged and auditable.
• Alarms & Observability — Define threshold- or condition-based alarm rules on any live telemetry stream. The observer system evaluates incoming device data continuously and fires notifications when conditions are met, giving you real-time visibility into the health of every machine and group in your project.
• Cloud Edge Fusion — Provision cloud devices — virtual machines running in the same cloud infrastructure as IronFlock (on Google Cloud, AWS, or equivalent) — that join your project alongside your physical edge devices. A cloud device participates in the same message routing context as any physical device, enabling direct communication between cloud-hosted compute and on-premises hardware as if both were on the same local network. This is available in the IronFlock cloud and can be enabled in virtual private cloud or on-premises installations upon request.
• App Distribution & Marketplace — Publish industrial apps publicly or privately. Set pricing, track usage metering, and receive automated payouts — turning operational software into a scalable product.

Multi-Agent AI for Industrial Operations

Traditional industrial AI projects are monolithic — one model, one purpose, one long integration cycle. IronFlock takes a different approach: multiple specialized AI agents, each scoped to a specific task, working together within a shared operational context.

An app for predictive maintenance might include one agent that monitors vibration patterns, another that queries historical failure data, and a third that generates maintenance work orders. A quality inspection app might pair a vision AI agent with a traceability agent that links defects to specific batches and process parameters.

These agents are defined declaratively in a lightweight YAML template, connected to live machine data, and orchestrated through a single conversation interface. Operators ask questions in plain language — the system routes to the right agent, pulls from the right data source, and returns actionable answers.

This is not a chatbot bolted onto a dashboard. It is AI that understands your machines, your processes, and your operational context.

Industrial Use Cases

IronFlock is built for teams deploying real operational software:

Beyond the Factory Floor

While IronFlock has deep roots in industrial automation, the platform is not limited to factory use cases. Its architecture — edge agents, containerized apps, real-time data, secure tunneling, and AI orchestration — applies wherever distributed devices need to be managed, monitored, and made intelligent:

• Building Automation & Monitoring — HVAC control, energy metering, occupancy tracking, and predictive maintenance for commercial and residential buildings
• Vehicle Fleet Management — Real-time GPS tracking, driver behavior analytics, maintenance scheduling, and route optimization across vehicle fleets
• Video Management & Surveillance — Edge-based video analytics, camera fleet management, AI-powered object detection, and secure remote access to live feeds
• Smart City Infrastructure — Environmental sensor networks, air quality monitoring, EV charging station management, smart lighting, waste management, and parking systems

Any scenario that involves a fleet of connected devices running software at the edge is a fit for IronFlock.

Cloud or On-Premises

IronFlock runs in the cloud or entirely on your own infrastructure. For sensitive industrial environments that require private networks, air-gapped deployments, or data sovereignty, the full system — including the AI runtime, data layer, and app delivery — can be deployed on-premises or in a private cloud. See On-Premises Deployment for details.

Quick Links