UAO at the ARC EIF 2025: Projects, Use Cases, and Products

Keywords: UniversalAutomation.org (UAO), ARC EIF 2025, IEC 61499 standard

Overview

UniversalAutomation.org (UAO) is an independent, non-profit association that promotes an open, hardware-independent software ecosystem aimed at transforming industrial automation. Its core mission is to decouple automation application software from the hardware it runs on, enabling true interoperability, portability, and innovation across vendors and platforms.

The core technology that powers the UAO’s vision of open, interoperable industrial automation is the UAO Runtime Execution Engine. It enables software applications to run independently of the underlying hardware and automation logic to be distributed across multiple devices in a network. The UAO Runtime is based on the IEC 61499 standard, which defines a component-based, event-driven model for industrial automation.

Key Features Include:

• Function Blocks (FBs): Modular software components that encapsulate control logic and can be reused across systems.
• Hardware Independence: Applications are written once and can be deployed on any compliant hardware.
• Distributed Control: Supports scalable, decentralized architectures ideal for Industry 4.0 and edge computing.
• Plug-and-Produce: Enables dynamic reconfiguration and integration of devices without reprogramming.

Key Benefits Include:

• Portability: Software can be reused across different hardware platforms, reducing vendor lock-in.
• Interoperability: Devices from different manufacturers can work together seamlessly.
• Lifecycle Efficiency: Applications can be incrementally improved and maintained without full rewrites.
• Cost Reduction: Less engineering effort, faster commissioning, and reduced downtime.
• Future-Proofing: Software remains usable even when hardware becomes obsolete.

The non-profit association is supported by a growing global community of OEMs, system integrators, end users, technology vendors and academics. All of these are collaborating to develop and maintain the shared UAO Runtime and software libraries.

IEC 61499 Standard

IEC 61499 is an international standard for distributed industrial automation systems. It defines a component-based, event-driven architecture for designing and deploying control applications across multiple devices and platforms.

Key Concepts Include:

• Function Blocks (FBs) are the core building blocks of IEC 61499, that encapsulate both data and behavior (logic). They promote modularity, reusability, and encapsulation and can be basic, composite, or service interface blocks.
• Event-Driven Execution, meaning that IEC 61499 uses event triggers to execute logic unlike traditional cyclic scan models, allowing for asynchronous, reactive, and deterministic behavior.
• Distributed Control allows applications to be distributed across multiple devices, with each device able to host multiple function blocks that communicate via events and data. This makes it highly suitable for edge computing, IIoT, and Industry 4.0 scenarios.
• Separation of Concerns. IEC 61499 makes a clear distinction between application logic, communication, and hardware configuration, thus enabling hardware-independent software development.

The architecture consists of applications built from networks of function blocks. Devices host resources, which execute these function blocks. The communication between blocks can be local or over a network. Management commands allow dynamic reconfiguration (e.g., deploying, starting, stopping blocks).

The hardware independence of the standard enables faster innovation cycles by allowing software developers to focus on functionality rather than hardware constraints. It enables simplified multi-vendor environments, reducing the complexity of maintaining diverse systems. Further, the event-driven nature supports the enhanced integration with IT systems, supporting modern technologies like AI, cloud, and digital twins. Use cases can be found in smart manufacturing, edge AI deployment, energy systems, water treatment, process automation and digital twins.

Barbara UAO Project with Acciona Smart Plant

David Purón, CEO of Barbara, presented their UAO project Acciona Smart Plant using their edge management and orchestration tool. The Spanish company Acciona Agua operates in over 25 countries and delivers solutions for desalination, wastewater and drinking water treatment, and full water cycle management for cities and industries. Their projects emphasize low-impact water treatment, using renewable energy sources and digital water management tools.

For their Acciona Smart Plant project they were looking for a centralized system to monitor and manage over 150 water treatment plants. They needed real time data instead of lab results every 24 hours. They put an emphasis on cybersecurity, data privacy and intellectual property (IP). They also wanted to introduce advanced data applications, including machine learning (ML) models in distributed environments.

Project challenges included the avoidance of hardware dependency resulting in a need to focus on software-driven solutions, the heterogeneity of plants, the limited connectivity from OT to IT, regulatory compliance with e.g., NIST and cybersecurity policies like IEC-62443 and the need for scalability to hundreds of plants. To make the project a success it was critical to ensure long-term support and maintenance.

Barbara’s platform enables distributed AI model deployment across heterogeneous OT environments using a software-defined infrastructure. Key components include:

• Barbara edge node, which acts as the local execution environment for AI models and control logic.
• Edge management & orchestration platform, which delivers centralized control for deploying, updating, and monitoring edge applications.
• Usage of MQTT PUB/SUB as communication protocols for lightweight, real-time data exchange between nodes and central systems.

Security layer deployed includes VPN & DMZ secure communication channels, a firewall to protects edge nodes and internal networks, while IEC-62443 compliance ensures industrial-grade cybersecurity.

SoftdPAC Runtime, a containerized version of the runtime execution engine shared by UniversalAutomation.org, enables deployment of virtual PLC logic at the edge, reducing reliance on proprietary hardware. AI models are deployed for predictive analytics, e.g., real-time chemical usage prediction, optimizing operational efficiency and reducing costs.

The data lifecycle includes acquisition from sensors, actuators, and PLCs, preprocessing by filtering and event handling at the edge, execution through AI inference and control logic run locally and the feedback loop through which results are fed back into the system for adaptive control.

The hardware-agnostic design supports multi-OEM environments with varying protocols and hardware, virtualized PLCs using IEC-61499 standard and avoids vendor lock-in as applications are portable across different hardware platforms.

In this project, Barbara’s Edge platform manages and deploys multiple IPC devices in the field allowing for $250,000+ annual savings per plant from optimized chemical usage. It also achieved a 74 percent reduction in Total Cost of Ownership (TCO) due to software-defined infrastructure and reuse of edge hardware. Barbara's solution includes the software infrastructure to deploy, orchestrate and maintain AI models on Schneider Electric P6 IPC devices across multiple sites as well while also supporting SoftdPAC runtime for deploying PLC logic at the edge—reusing the same infrastructure.

The UAO Runtime Execution Engine was the enabler for this project. Thanks to the vendor-neutral, open ecosystem, control applications can be developed once and deployed across heterogeneous plants, eliminating costly rewrites and reducing vendor lock in. By leveraging the key features of the technology, Barbara ensures a robust, secure, and standardized application development, with a minimal time to market. This solution, built on key concepts of UAO, supports a long-term vision that spans from basic data acquisition to intelligent, automated operations with minimal resource requirements. Furthermore, the use of UAO’s hardware-agnostic runtime execution engine protects the investment from hardware obsolescence, allowing seamless migration and reuse of software - even when edge devices are replaced.

Flexbridge-LTU AI Tools

Midhun Xavier of the Luleå University of Technology (LTU) and Valeriy Vyatkin, CEO of Flexbridge presented the Flexbridge-LTU project, a forward-looking initiative that integrates AI technologies, particularly Large Language Models (LLMs), into the engineering of IEC 61499-based distributed control systems. The project addresses the growing complexity of industrial automation by enabling natural language-driven development, reducing manual coding, and promoting modular, reusable automation components.

At the heart of the project is the Function Block Assistant (FBAssistant), an AI-powered tool that interprets natural language specifications and automatically generates IEC 61499 function blocks. This assistant is embedded within an engineering workflow that includes:

• IEC 61499 IDE: For visual programming and integration.
• Function Block Repository: For storing and reusing generated blocks.
• Soft PLC and Simulation Tools: For validating logic before deployment.
• State Machine Generation: For modeling system behavior and transitions.

This workflow allows engineers to move from high-level requirements to executable logic with minimal manual intervention, significantly accelerating the development cycle.

The project introduces a skills-based architecture, where intelligent devices expose their capabilities as skills—modular services implemented as IEC 61499 function blocks. These skills can be orchestrated into complex applications using AI agents. They are designed to be interoperable and reusable, and they can be accessed via standard industrial communication protocols such as HMI, OPC UA, and HTTP.

An AI Agent complements the FBAssistant by handling the orchestration of skills into higher-level processes. It supports the simulation of workflows using state machines, the validation of logic before deployment and the dynamic composition of services based on system goals. This agent enables Manufacturing-as-a-Service (MaaS) by allowing flexible, on-demand configuration of production processes.

The use case presented was an integrated assembly solution (IAS), that demonstrates how multiple skills can be orchestrated to perform a coordinated manufacturing task. It showcases service-based interaction between devices, AI-assisted planning and execution and scalability through modular design. This use case exemplifies how the system supports decentralized, secure data exchange, aligning with goals of sustainability and circular manufacturing.

Engineering benefits include a reduced engineering time through automation of function block creation, improved maintainability via modular, skill-based design, enhanced flexibility in system configuration and orchestration as well as support for Industry 4.0 paradigms, including MaaS and digital twins.

Cloud Virtual Commissioning - CloViC

Mikhail Kolesnikov of the Aalto University, Helsinki and Valeriy Vyatkin, Professor at Aalto University and CEO of Flexbridge, presented the online web platform cloud virtual commissioning platform CloViC. It is enabling remote and collaborative commissioning for system integrators that aims at reducing the time, cost, and energy spent on commissioning intelligent automation systems. The platform leverages standards such as IEC 61499 and IEC 61131-3 to support both greenfield and brownfield automation projects. The platform aims to reduce commissioning costs, improve system reliability, and support remote collaboration.

CloViC’s technical capabilities include:

• IEC 61499 Integration: Supports event-driven, modular, and distributed control logic.
• IEC 61131-3 Runtime in the Cloud: Enables simulation and testing of legacy PLC logic.
• Hybrid System Support: Allows integration of both IEC 61499 and IEC 61131-3 PLCs in a single virtual environment.
• Networking of PLCs: Simulates real-world industrial networks for comprehensive testing.

The automation workflow starts with performance modeling, which simulates system behavior under various operational conditions, followed by automated testing to reduce manual testing effort and increases reliability. The next step is an AI-based analysis which applies machine learning to interpret test results and detect anomalies. The workflow concludes with the application of reusable control logic using libraries of tested control fragments to accelerate development and reduce errors.

Case Studies for CloViC

Case Study 1: Discrete Manufacturing (Greenfield)

• Scale: 50–100 PLCs
• Savings: €525,000 (30% reduction in commissioning costs)
• Methods: Performance modeling, automated testing, reuse of control logic

Case Study 2: Oil, Gas, Water Treatment (Brownfield)

• Scale: 5–10 PLCs per venue
• Challenges: Legacy hardware, supply chain issues
• Savings: 20 percent reduction in fixed costs, 50 percent reduction in per-venue costs
• Methods: Gradual hardware replacement, reuse of logic, automated testing

Case Study 3: System Integrator

• Scale: 15–20 projects/year, 4 global teams
• Savings: Up to $78,000 per project; $1.5M annually
• Methods: Remote collaboration, version control, shared environments, virtual commissioning

Right now, the CloViC web platform is in the phase of pilot access and available for industry partners. The proof of concept includes performance benchmarking and automated testing scenarios. Academic validation has been achieved through extensive testing on student projects to ensure scalability and usability.

Strategic benefits of CloVic include the reduction of commissioning time and cost, the minimization of on-site work and travel, the support of both legacy and modern automation systems, all while it enables global collaboration and remote engineering.

Aimirim’s Smart Boiler Optimization with UAO

Renato Pacheco Silva, CEO of the Brazilian startup Aimirim presented their use case for UAO in a smart boiler optimization. The startup company develops different technologies applied to industry, such as observability, simulation, virtualization, artificial intelligence, machine learning and others.

The goal of the project was to enhance steam quality and reduce fuel consumption in industrial boilers through advanced control and real-time optimization technologies integrated with existing supervisory systems.

The first core technologies applied was the advanced control OPPER Arandu MPC/MFAC, offering real-time adaptive control that learns fuzzy rules online using artificial intelligence algorithms. OPPER Arandu improves the control performance over time and maintains satisfactory control  on different process conditions. It directly influences the fuel injection valves to stabilize steam pressure and flow while increasing equipment performance. The second technology used was the real-time optimization tool OPPER Arandu RTO, which optimizes performance across multiple boilers using efficiency curves and process data. It operates in parallel with existing PLC and PID systems  to ensure smooth operations.

The system architecture consists of three layers. The hardware layer consists of an edge device running IEC61499 runtime  connected to the existing PLC and supervisory system. Integration is smooth  and advanced control can be activated/deactivated via the supervisory interface. The data flows in real-time from sensors connected to a PLC, into Aimirim’s control logic in an edge device and goes back to the actuator through PLC communication.

The implementation was divided into two major parts: data analytics pipeline and scenario identification.

The data analysis pipeline part started with data collection, executed in 2024, followed by data cleaning through filtering, unit conversion, and outlier removal. The next steps were the feature engineering using thermodynamic calculations and latent variable extraction and clustering through the identification of operational regimes using key process variables.

The second part concerned the key variables for scenario identification. The scenario was identified using boiler variables like steam pressure and temperature, boiler furnace and exhaust gas temperatures, fuel supply speed and weight, water supply flow and temperature, as well as steam flow and setpoints.

Clustering operational data through techniques to deal with transients and autocorrelated time series identified some important clusters:

• Cluster 0 (78.54 percent): Operational scenario with high steam flow, high exhaust temperature – represents standard production.
• Cluster 1 (20.20 percent): Mid-range operation scenario.
• Cluster -1 (1.26 percent): Outliers – excluded from analysis.

With these, automatically identified scenarios, steam pressure variability reduction of 36.8 percent were observed, resulting in a fuel consumption reduction of 4.21 percent, while demonstrating improved control, shown by a more stable dosing screw velocity and steam mass flow rate under OPPER-enabled conditions. Renato presented the following deployment roadmap(s). The first relies on ideal conditions, like Ethernet TCP/IP access to PLC, common PLC vendors (e.g., Siemens, Rockwell), presence of a switch port and automation technician, and availability of historical process data. In this case the timeline looks like this:

• Assessment: 1–3 weeks
• Data Analysis & Setup: 2–5 weeks
• Commissioning: 1–2 weeks
• Go Live & Training: 1 week

The second roadmap takes non-ideal conditions into account. These are the absence of Ethernet access or the usage of uncommon PLC vendor, without  historical data or if in-house automation technicians are not available. As this situation requires on-site assessment, external partners, and IT coordination, the execution will require additional time of two to four extra weeks depending on the constraints.

The solution is designed to scale across multiple plants without replacing  current automation assets with edge computing in the supervision layer,  defining the behavior of the plant using software defined architecture, leveraging historical data, clustering models, advanced control and providing remote support and diagnostics.

Gr3n – Microwave Assisted Depolymerization

Franco Antonio Cavadini, CTO of Gr3n, presented the advancements achieved in the MADE/MODUS project. Gr3n SA is a Swiss company founded in 2011 that has developed an innovative chemical recycling process based on microwave technology. The process, known as Microwave Assisted Depolymerization (MADE), enables the depolymerization of Polyethylene Terephthalate (PET) into its monomeric components: Purified Terephthalic Acid (PTA) and Monoethylene Glycol (MEG). These monomers can then be reused to synthesize new PET with properties equivalent to virgin material.

The project focuses on scaling this PET chemical recycling process to an industrial level through the use of microwave depolymerization. One of the enabling technologies adopted in the project is the UniversalAutomation.org (UAO) runtime execution engine, based on the IEC 61499 standard. The project includes a demonstration plant (MADE) with a PET processing capacity of 20 kg/h. It features a small-scale industrial microwave reactor (1:5 scale), and a revamped software library for automation and control.

The MADE demonstration plant showcases Gr3n’s core process in a scaled-down industrial environment. It includes a second-generation microwave reactor, integrated MEG distillation, and a simplified PTA crystallization unit. The plant comprises 6 process cells, 40 vessels, 7 heat exchangers, 73 on-off valves, 6 control valves, 40 motors (28 with VSDs), 128 sensors, over 500 I/Os, and 9 control cabinets. The facility was built and commissioned in 18 months, following process industry standards for design and ISA-88 for control system design.

The automation architecture is structured into three layers:

1.    Software Stack:

• Based on IEC 61499 for modular and distributed control.
• Utilizes PostgreSQL and InfluxDB for data storage.
• Includes integrated predictive analytics and visualization tools.

2.    Execution Model:

• Event-driven, object-oriented, and component-based.
• Supports deterministic real-time control and asynchronous IT communication.
• Modular runtime with an OS abstraction layer.
• Compatible with Docker, Windows, Linux, and RTOS environments.

3.    Interfaces:

• South-bound: EtherCAT, Profinet, CANopen, OPC-UA, etc.
• North-bound: WebSocket, MQTT, AMQP, OPC-UA Server.
• East-bound: Distributed real-time control.
• West-bound: HMI and visualization.

The IEC 61499 implementation strategy includes control logic based on Basic Function Blocks (BFBs), allowing for multiple implementation patterns: a monolithic BFB, composite BFBs for modular design, and networks of BFBs for maximum reusability. Key benefits include hardware independence, scalable and upgradable architecture, seamless OT-IT integration, and support for polyglot persistence (SQL, NoSQL, time-series databases).

The proprietary equipment (PEQ) is central to Gr3n’s licensing model (a future full-scale plant will include 9 PEQs for a total productivity of 30,000 ton/year of PTA). It features a modular architecture focused on high-value components such as microwave generators. The design enables pre-assembly testing and internal quality control. Gr3n manages engineering and software development in-house, while excluding low-value equipment (e.g., vessels and pumps) from the proprietary scope.

The automation hardware architecture includes a centralized main cabinet for power supply and runtime control, remote I/O cabinets per reactor, and dedicated microwave cabinets housing the power supply unit (PSU) and integrated I/Os.

The updated software library introduces a process-driven architecture with self-similar object orientation, enabling hierarchical abstraction with well-defined interfaces. It separates concerns across initialization, field interface, OT-IT interface, and core logic layers, while supporting HMI integration and AI-readiness. Applied components include digital/analog I/Os; on/off, sealed, control, and solenoid valves; on/off and VSD motors; and magnetron-based microwave generators.

The strategic impact of Gr3n’s approach is evident in its licensing model: the PEQ is central to the business, protected by intellectual property and trade secrets. From a sustainability standpoint, the process achieves significant reductions in environmental impact: –29 percent CO₂ eq and –39 percent MJ of non-renewable energy use compared to virgin PET. Additionally, the modular equipment and software architecture ensure high scalability and rapid deployment across licensed plants.