Representative End User Clients
Representative Automation Clients
Representative Software Clients
Siemens, a Global Gold Sponsor of ARC Advisory Group’s 24th Annual ARC Industry Forum in Bengaluru, was represented by Vivek Roy, Head IT/OT Integration, Digital Industries, Siemens Limited. In his presentation, The OT Transformation Journey: Data Meets Defense, Roy examined how manufacturers can build the data integration and cybersecurity foundations needed to support industrial AI, scalable digitalization, and increasingly connected IT and OT environments.
As industrial organizations expand their use of AI, the quality, accessibility, and context of operational data become increasingly important. Machines, production systems, products, supply chains, and buildings generate large volumes of data, but much of it remains distributed across different systems and organizational silos.
Roy’s presentation focused on the architecture needed to turn this fragmented operational data into information that can support industrial AI, analytics, optimization, and more autonomous decision-making. At the same time, greater connectivity between IT and OT introduces new cybersecurity requirements, making secure data integration a critical part of the transformation.
Roy’s presentation can be viewed on YouTube or here:
Industrial AI Starts with Industrial Data
Manufacturing has progressed from manual production to automation and digitalization, with AI now creating opportunities for more adaptive and increasingly autonomous operations.
Siemens describes this progression as moving from automated production toward adaptive production and, eventually, future production models such as autonomous factories and the industrial metaverse. AI can support faster innovation, adaptive manufacturing, data-driven improvements, autonomous decision-making, and future-ready operations.
However, these capabilities depend on the data available to the AI systems.
Industrial organizations collect information from machines, production systems, products, supply chains, buildings, and other operational assets. The challenge is not simply collecting more data. Organizations must connect data that exists in different systems, make it usable across the lifecycle, and give it sufficient operational context.
Siemens describes the process through four broad stages:
Collect: Gather data from real-world operations.
Connect: Break down data silos and make relevant information accessible.
Contextualize: Give the data meaning through a common data fabric.
Act with confidence: Use contextualized information to support informed decisions and actions.
This progression is intended to transform raw industrial data into information that higher-level applications, analytics, and AI systems can use with greater confidence.
Moving Beyond the Traditional Automation Pyramid
Siemens illustrates how industrial data can progress from collection and connection to contextualization and informed decision-making
Traditional industrial architectures have often been represented as a hierarchical automation pyramid, beginning with sensors and actuators and progressing through PLCs, distributed control systems, SCADA, MES, PLM, and ERP systems.
That model becomes more difficult to maintain as manufacturers seek to connect a growing number of applications, cloud services, edge systems, enterprise platforms, and AI tools.
Rather than relying on isolated point-to-point connections between systems, Siemens emphasizes the importance of a flexible data integration layer between shop-floor OT and higher-level IT systems.
Roy noted that IT and OT increasingly need to work together to unlock the value of operational data. He described this relationship as collaboration rather than simply convergence because the two environments retain different requirements, responsibilities, and operating priorities.
The data integration layer provides the foundation for that collaboration.
Building the Data Integration Layer
Siemens organizes the data integration layer around three primary capabilities:
Southbound connectivity to machines, controllers, sensors, and other OT data sources.
Data harmonization, contextualization, and management to make industrial data understandable and usable.
Northbound connectivity to enterprise applications, MES and MOM systems, ERP platforms, cloud environments, and other IT solutions.
The architecture can support both greenfield and brownfield environments and different levels of the manufacturing hierarchy, from individual machines and production lines to factory and enterprise systems.
Contextualization is particularly important. Time-series values alone may provide limited meaning to an application or AI system. That data becomes more useful when it is associated with the machine, production process, product, batch, operating condition, or other relevant industrial context.
A scalable data integration layer can therefore reduce the need for individual systems to build and maintain separate connections to each source of operational data.
Bringing Intelligence Closer to the Shop Floor
Industrial edge computing is becoming an important part of this architecture because not every industrial workload needs to run in the cloud.
Manufacturers may have requirements related to latency, data sovereignty, connectivity, cybersecurity, or operational continuity that make local processing preferable for certain applications. Edge infrastructure can allow organizations to process and contextualize operational data closer to where it is generated.
Siemens’ Industrial Edge architecture includes northbound and southbound connectivity, along with an Industrial Information Hub that can help manage and contextualize shop-floor information.
Applications running on the edge can include industrial AI, analytics and diagnostics, HMI and SCADA software, virtual PLCs, digital twin capabilities, and customer-developed software.
This approach also allows manufacturers to decide which information should remain within the operational environment and which data should be shared with enterprise or cloud applications.
Start with the Use Case
Technology architecture alone does not determine whether an IT/OT integration initiative will deliver value.
Roy emphasized that implementations should be driven by specific use cases. Without a clearly defined operational objective, organizations can continue adding technology and connectivity without establishing a clear measure of success.
Potential use cases include:
Track and trace.
Order execution.
Paperless production.
Production optimization.
Energy transparency.
Product carbon dioxide footprint information.
Quality management.
Maintenance.
Reporting.
MES and MOM integration.
Different use cases may require different implementation approaches. Depending on the application and operating environment, organizations can use edge computing, SCADA or HMI architectures, controllers, cloud connectivity, or a combination of these technologies.
The common requirement is a scalable method for connecting, managing, and contextualizing operational information.
Broker-based architectures can also help reduce point-to-point integration complexity. Siemens identifies Unified Namespace as one approach for making contextualized industrial data available to multiple applications through an enterprise message broker.
From Data to Defense
Greater connectivity also expands the cybersecurity requirements surrounding industrial data.
Traditional OT cybersecurity has frequently relied on cell protection, network segmentation, and defined security boundaries around machines or production areas. As data moves more dynamically between OT systems, enterprise environments, edge infrastructure, and cloud applications, organizations require more granular mechanisms for controlling communication.
Siemens positions Zero Trust Network Access as an extension of traditional cell protection for increasingly connected OT environments.
Siemens extends traditional OT cell protection with zero trust, using zones, conduits, and microsegmentation to provide more granular access control
The approach aligns with the zones-and-conduits model associated with IEC 62443. Zones can represent individual machines, users, applications, or legacy networks, while conduits control permitted communication between them.
Zero trust can further reduce network zones to individual hosts and applications through microsegmentation. Policies can then determine which identities are allowed to access specific services rather than relying entirely on network location or IP addresses.
Applying Zero Trust in OT
Roy illustrated how zero trust can be applied to communication between SCADA systems and PLCs while continuing to use existing industrial protocols such as S7 communication and OPC UA.
In the Siemens example, SINEC Secure Connect components establish controlled communication paths between systems. Policies based on identities and attributes determine access to specific services.
This architecture can help organizations implement IEC 62443 zones and conduits while maintaining greater flexibility as machines, applications, or clients move within the industrial environment.
The objective is not to eliminate established OT network structures. Instead, zero trust can add more granular access controls to existing cell protection and segmentation strategies.
Legacy environments can also remain part of the architecture through gateways that control communication between older networks and other security zones.
From Point-to-Point Complexity to Scalable Integration
The growing use of industrial AI makes operational data increasingly valuable, but simply connecting more systems does not guarantee better outcomes.
Manufacturers need architectures that can collect, integrate, contextualize, and securely distribute industrial information according to clearly defined operational requirements.
Roy summarized the approach through a straightforward sequence:
Use Cases → Architecture → Technology → Deployment
Starting with the use case helps organizations define the required architecture before selecting technologies and determining how they should be deployed.
As manufacturers move toward more adaptive, data-driven, and autonomous operations, the ability to connect IT and OT securely will become increasingly important. A flexible data integration layer, supported by industrial edge computing, contextualized information, scalable integration approaches, and appropriate cybersecurity controls, can provide the foundation needed to turn industrial data into actionable operational intelligence.