Digital Twins How Atlas Copco Links Production and Field Data Through MATLAB Models

Related Vendors

Mathworks Deutschland

FAULHABER Drive Systems

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Wuxi InfiMotion Technology Co., Ltd.

With more than 250,000 connected machines in operation, Atlas Copco is building a digital ecosystem with real-time data collection that enables data-driven decision-making, operational efficiency, product personalization, and increased customer value.

Digital twins at Atlas Copco go beyond traditional simulation models. They are dynamic, highly accurate digital replicas of physical products that incorporate data from the design, production, operation, and service phases. Developed with Matlab and Simulink, these digital twins model the interactions between physical systems, sensors, and controllers, enabling precise replication of real scenarios and efficient testing that surpasses the limitations of traditional testing approaches.

The digital twin of Atlas Copco creates a single source of truth for the entire value chain by consolidating millions of real-time data points from production, testing, and design systems. In this way, the company promotes continuous improvements through evidence-based and data-driven decisions.

The digital twin framework for compressors at Atlas Copco consists of four key phases (Fig. 1): In the "As-Designed" phase, the focus is on prototyping and feasibility analyses. In the "As-Configured" phase, the products are adapted to exact customer requirements. The "As-Built" phase includes the collection of production data and ensuring traceability, while in the "As-Maintained" phase, operational data is used for predictive maintenance and continuous optimization.

Each of these phases contributes to a holistic approach that ensures continuity from the concept to the design and construction of compressors to their operation. This makes the digital twin a dynamic, ever-evolving resource that supports decision-making at all levels.

To promote the digital twin strategy, Atlas Copco established the Model-Based Engineering (MBE) Community of Excellence—a global network that connects teams from different departments and areas. This community uses Model-Based Design (MBD) principles to share knowledge, standardize processes, and foster collaboration. It plays a key role in aligning design, configuration, manufacturing, and maintenance, ensuring that insights gained in one area are quickly disseminated across the entire company.

A key challenge lies in balancing model accuracy with development efficiency. The dynamic behavior of the product must be precisely captured for the development of control algorithms, without unnecessarily consuming development time and computing resources. This is a balancing act between simulation accuracy and performance that Atlas Copco continuously refines through testing, collaboration, and data.

The GA VSDs (Variable Speed Drive) oil-injected screw compressor series illustrates this approach. It is the first product to be fully developed using digital twin technology, enabling optimized control algorithms, CI/CD workflows, more efficient operation, and services such as predictive maintenance.

Infrastructure for Digital Twins

A robust and scalable MBE platform forms the operational backbone of Atlas Copco's digital twin ecosystem. At its core is the Matlab Production Server (MPS), which serves as a bridge between operational technology (OT) and information technology (IT) systems. The platform connects physical systems, digital models, and enterprise-wide software tools, enabling engineering teams, marketing analysts, and service managers to access and interact with the same datasets and insights. Additionally, MPS allows models and applications to be shared with non-Matlab users, thereby accelerating development processes and collaboration.

The MBE platform provides three core functions:

• Central Processing: Matlab Production Server executes models centrally, ensuring security, consistency, scalability, and low-latency processing of the provided code.
• API-based integration: The platform integrates with tools such as Power BI, SAP, Matlab Web App
• Server, and many other applications, easily bridging the gap between operational technology (OT) and information technology (IT).
• Cloud scalability: The cloud infrastructure enables dynamic scaling to meet process requirements, from overnight model testing to global web applications.

The MBE infrastructure evolved by integrating existing components and systems such as PLM, ERP, IoT, and test data. Digital twins are centrally provided and accessible to relevant teams, enabling cross-functional insights. By connecting data silos, teams, and tools, digital twins have been scaled into an enterprise-wide solution, offering benefits such as improved transparency, automation, and performance optimization across all phases of the product lifecycle.

Configuration And Sales

The "As Configured" digital twin optimizes the sales process through technically sound and validated recommendations. These are based on a quick simulation of compressor performance for specific customer requirements (Fig. 5), such as altitude, ambient temperature, humidity, and usage patterns.

Production And Quality

In the "As-Built" phase, digital twins bridge the gap between production lines and quality systems. Atlas Copco integrates machine tool data, measurement systems, and test validations into a unified quality framework. Data from sensors, machines, and operators are linked to the digital model of each unit, ensuring seamless traceability. The in-house design of all key compressor components enables the underlying digital fingerprinting, supporting root cause analysis, production optimization, and compliance with regulatory requirements.

Operation And Maintenance

The "As Maintained" digital twin enables predictive maintenance through the analysis of telemetry data from more than 250,000 connected machines. Atlas Copco's SmartLink platform processes real usage patterns, detects anomalies, and recommends actions. By integrating SmartLink data with design and production records, the system enables full traceability and allows for root cause analysis of machine behavior during operation. This holistic approach makes it possible for compressors to proactively alert service teams to emerging issues, helping to reduce downtime, extend equipment lifespan, and enhance operational reliability.

Beyond the Product

The digital twin strategy of the Swedish company goes beyond products and also includes systems used for testing and validation. Advanced test cells create an environment for real-time modeling, simulation, algorithm prototyping, and testing under conditions that closely resemble real-world scenarios (Fig. 5). Each test cell is equipped with Speedgoat controllers running Simulink- and Simscape-based digital twins in real time. These control systems dynamically simulate complex load conditions, environmental fluctuations, and machine behavior.

Moreover, control algorithms can be spontaneously implemented and adjusted. With the new MBE infrastructure, software and hardware can be co-developed in an agile, iterative cycle, enabling rapid prototyping of intelligent control functions such as energy-saving modes or advanced fault detection.

Unified communication protocols like OPC UA, centralized data storage, and reusable model components ensure that insights from one product or test can be transferred to others. By storing and organizing extensive test data, Atlas Copco is building a valuable knowledge base for continuous innovation. This approach, initially tested in the GA VSD product line, is now being scaled company-wide, integrating insights into both products and their development processes.

Conclusion: Current Status And Outlook

Atlas Copco is continuously enhancing its digital twin capabilities, although some technical challenges and development goals remain. A key task is the seamless integration of Matlab-based models with programming languages like Python and platforms like Databricks to improve flexibility and interoperability between teams. Another challenge is deploying digital twins on edge devices to enable real-time insights and predictive control. To address these challenges, Atlas Copco is developing simplified, faster versions of its highly accurate, computationally intensive digital twins, known as surrogate models.

Furthermore, Atlas Copco is preparing to introduce intelligent test cells across all business areas. The goal is to establish real-time simulations and tests as a standard throughout the entire product lifecycle.

The close collaboration with MathWorks and the partnership with training, consulting, and development teams aim to refine the frameworks for control systems and the integration of AI-driven workflows with high-precision models. Through these initiatives, Atlas Copco seeks to enhance interdisciplinary collaboration, better align AI with physical modeling, and enable virtual system simulations without physical prototypes. The goal is to transform data-rich environments into insight-rich ecosystems and support continuous improvement through digital twin technology.

*About the authors:
Dr. Simone Giorgetti is Manager of Model-Based Engineering at Atlas Copco,
Dr. Dmitry Samarkanov is Technical Expert at MathWorks, and
Dr. Mo Anas is Regional Engineering Manager at MathWorks.

(mc)

Smart-Factory

Access solutions for the Smart Factory - Industry 4.0