👁️
AI Observes
IRIS captures physical reality with CNN for vision, IoT for process signals, and direct connections with personnel through conversational apps. It sees what humans can't see in real-time.
Research & Intelligence
The IRIS Observatory
We monitor quarterly the deployment of IRIS systems across all industries and countries in the West. Because if an industrial AI is IRIS, we know it's driving the 3 core values the European Initiative defends.
What is an IRIS System?
IRIS (Industrial Reality Intelligence Systems) defines a new category of industrial AI software designed to understand shop floor reality in real-time and convert it into actionable decisions — always with humans at the center of operations.
"IRIS is an industrial reality intelligence system that observes the factory in real-time, integrates its critical data, interprets it with advanced AI, and converts it into recommendations and concrete actions — but it's always the human who DIRECTS operations."
How does IRIS capture factory reality?
- Convolutional Neural Networks (CNN): Computer vision to detect defects, verify operations, and track processes in real-time.
- Human-AI communication apps: Conversational interfaces where shop floor personnel talk directly with all the AIs in the system.
- AI-interpreted IoT systems: Sensors, PLCs, SCADA, and process signals feeding anomaly detection models.
- Central GPTs: Generative AI systems that accumulate all information, interpret it, and predict what will happen on the plant.
Why Does the European Initiative Promote IRIS?
European Initiative Industry 4.0 promotes that companies deploy IRIS systems, that tech companies develop IRIS solutions, that training organizations build IRIS methodology curricula, and that professionals get IRIS certified. The reason is clear:
At the center of an IRIS system is always the human being.
An AI software within the IRIS category always has at its core philosophy that AI captures factory reality and puts all that information in the hands of plant personnel so that the human always DIRECTS operations.
This architecture guarantees highly productive factories that will always have the "human in the loop" — building operations that empower workers instead of replacing them.
Because IRIS is so aligned with the 3 pillars of the European Initiative 4.0 mission, the institution conducts quarterly tracking of this human-empowerment AI deployment across all industries and countries in the West — driving massive adoption of these systems throughout all our operations for the enormous benefits this will bring after the AI transformation of the entire business fabric.
Pillar 1
IRIS Guarantees Human-Empowerment AI
By design, an IRIS system never replaces the human. AI observes, interprets, suggests, and coordinates — but decision-making and final action capability always remains in human hands.
🧠
AI Interprets
Central GPTs accumulate all information, correlate it with context (shift, product, equipment) and predict what will happen. It understands why things happen.
💡
AI Suggests
IRIS generates actionable recommendations in natural language: what to adjust, when to stop, what to review. It proposes, but doesn't impose.
👨🔧
Humans Decide
Operators, technicians, and managers receive information and make decisions. The human always DIRECTS operations — empowered by AI, never replaced.
"IRIS is designed to enhance the capability of operators, technicians, and managers — standardize best practices and facilitate decision-making — avoiding the 'black box that decides alone and displaces humans' model."
Pillar 2
IRIS Guarantees Safe and Controllable AI
Security requirements for an IRIS system are extremely high. They're designed to harness all AI power while guaranteeing by design that AI will never get out of control in the factory.
Guiding principle: Separate "thinking" (AI) from "having hands" (OT action).
AI can talk; the hands remain with the human.
Ring 1
External & Agile Zone
A zone designed for agility: secure remote access to information, analysis, reporting, support, collaborative work, and coordination functions that don't involve "hands" on OT.
✓ Allows
- Reports and operational analytics
- User and role management
- Executive visualization
- Support and diagnostics
✗ Doesn't allow
- Acting on PLCs or control
- Access to OT networks
- Modifying machine parameters
"Agility outside, control inside."
Ring 2
DMZ + Gateway
An explicit boundary: everything crossing between zones passes through a DMZ and a gateway with minimal surface. The goal isn't "connecting" — it's filtering, logging, and controlling.
Non-negotiable principles:
- One door well-defined
- Explicit policy: only what's permitted
- Complete logging: every session, every attempt
- Segmentation: border control, not a free bridge
Typically contains:
- Edge integration gateways
- Access controllers
- Proxies and inspection
- Complete logging
"One well-guarded door is worth more than a hundred half-closed ones."
Ring 3
Human-Operated OT (No AI)
The key IRIS security point: The most critical zone — OT network and systems operable ONLY by humans. Here lives the tactile "control plane" where there is NO AI whatsoever and where only a human can authorize any intervention.
IRIS can feed information, alerts, opportunities, and recommendations.
But EXECUTION within the most critical ring is done through human-operable software, with permissions, procedures, and traceability.
Why does this change the game?
- Avoids the fear argument: "What if the AI goes crazy?"
- Answers the corporate requirement: "How do you guarantee no unauthorized actions in OT?"
- Aligns deployment with human governance, auditing, and certification.
"AI doesn't command in OT: OT is governed with human hands."
"The day your architecture separates thinking from having hands, AI stops being a risk and becomes infrastructure."
The Role of the IRIS Observatory
The IRIS Observatory is the strategic and technical body of European Initiative Industry 4.0. Its main mission is to conduct quarterly tracking of IRIS system deployment across different industries and countries in the West.
Why do we monitor IRIS deployment?
If an industrial AI system is IRIS, we know that AI is driving the 3 core values the Initiative defends:
- Human-Empowerment AI: Humans always direct, AI empowers their capabilities.
- Safety AI: 3-ring security architecture guaranteeing safe and controllable AI.
- Western Reindustrialization: With human-empowerment AI at the core of transformation.
That's why we drive massive IRIS system adoption in industry throughout all our operations, for the enormous benefits this will bring to the West after the transformation AI will produce across the entire business fabric.
📊
Quarterly Reports
We publish reports on IRIS technology absorption rates by sector and country. State of ethical industrial AI across the West.
⚠️
Safety AI Alerts
We detect and communicate risks in industrial AI architectures. Technical notes on secure vs insecure implementations.
👥
Employment Impact
Human-side tracking: new roles, reskilling, fair transition cases. We document real worker empowerment.
🏛️
Think Tank
Roundtables with industrialists, experts, unions, and regulators. Policy recommendations for governments.
"IRIS is not about replacing people: it's about giving them systems to decide with the truth on the shop floor."
— Sebastián J. Brau, author of the IRIS Playbook and director of the IRIS Observatory's methodological framework, is one of the world's most recognized experts in industrial AI implementations.
Technical Keys of an IRIS System
An IRIS system isn't simply "AI in the factory." It's a complete architecture with specific components that guarantee human empowerment.
📷 Observed Reality, Not Just Data
IRIS starts from physical reality:
- Computer vision (CNN): Defects, component presence/absence, work sequences, real times
- Process signals: Temperatures, pressures, consumption, vibrations, alarms, machine states
What's really happening on the line right now?
🗂️ Semantic Factory Model
IRIS builds a semantic model of the production system:
- Machines → Cells → Lines → Value Streams
- Product families, variants, references
- Standard operations, SMED, inspections
- Human roles and responsibilities
Level change: From "Tag T104 = 89°C" to "The paint booth is operating above normal range; defect risk exists."
💬 Conversational Interfaces
IRIS connects humans with AI through:
- Direct communication apps: Chat and voice to query IRIS
- Contextual alerts: By business impact, not just threshold
- Role-based interfaces: Operator, maintenance, quality, management
Shop floor personnel talk directly with all the AIs in the system.
🧠 Predictive Central GPTs
IRIS integrates generative AI that:
- Accumulates all plant information
- Correlates with historical context
- Predicts what will happen
- Generates natural language explanations
Humans receive predictions and explanations to make better decisions.
🔄 Living Memory & Learning
Every event feeds the IRIS living memory:
- What happened and in what context
- What action the human took
- What result it had
Over time, IRIS learns what works best — but humans always validate.
Stay Informed with Observatory Reports
Subscribe to receive our quarterly IRIS reports, Safety AI alerts, and employment impact studies. Be part of the community defining ethical industrial AI.
Observatory FAQs
What is an IRIS system?
IRIS (Industrial Reality Intelligence Systems) is a new category of industrial AI software designed to understand shop floor reality in real-time, with humans always at the center of operations. AI observes, interprets, and suggests — but the human always DIRECTS operations. An IRIS system integrates CNN for vision, AI-interpreted IoT, human-AI conversational interfaces, and central GPTs that predict what will happen.
Why does the European Initiative promote IRIS?
Because IRIS is perfectly aligned with our mission's 3 pillars: (1) Human-empowerment AI — humans always at the center, (2) Safety AI — 3-ring security architecture guaranteeing controllable AI, and (3) Western Reindustrialization with human-empowerment AI. When we monitor IRIS deployment, we know that AI is driving our values.
What is the 3-ring security architecture?
It's the guiding IRIS security principle: separating "thinking" (AI) from "having hands" (OT action). Ring 1: External agile zone for analytics and reporting. Ring 2: DMZ + gateway with strict border control. Ring 3: Human-operated OT network where there is NO AI — only a human can authorize any intervention on machinery.
How often does the Observatory publish reports?
The Observatory publishes quarterly reports on the state of IRIS system deployment in the West, by sector and country. Additionally, it issues Safety AI alerts when risks are detected and publishes employment impact studies documenting real worker empowerment.
How does IRIS guarantee AI won't replace humans?
By architectural design. In an IRIS system, AI can observe, interpret, predict, and suggest — but executing any action on OT (plant machinery) requires human validation. Security Ring 3 is an AI-free zone where only humans operate. This guarantees permanent "human in the loop."