RTA's Blog
The IT/OT Network Divide
Jun
Information Technology (IT) and Operational Technology (OT) were once entirely separate worlds, but smart manufacturing and plant floor digitization required these networks to connect. However, bridging the gap between IT and OT networks remains a primary challenge for control engineers, system integrators and corporate IT teams due to fundamental differences in each system.
What is Information Technology (IT) vs. Operational Technology (OT)?
Information Technology (IT) comprises the collection of assets and technologies that allow the business of a company to be transacted.
Operational Technology (OT), on the other hand, comprises the processes, equipment and materials that produce products. Those products can be electricity, water, automobiles or any other of a million different goods.
The Challenge of IT and OT Convergence
The worlds of OT and IT coming together have been difficult.
Many years ago, OT used standalone processes for manufactured products. With widespread adoption of Industry 4.0, it has become increasingly dependent on networking technology to closely coordinate the operation of processes and equipment and uses Cloud applications to increase efficiencies, quality and speed. The worlds of OT and IT are coming together, and it has been difficult.
One of the main factors making it difficult for IT and OT staff to work together is that each has different goals, technologies and operating philosophies. Each team has very little knowledge of the other team and difficulty understanding what they do and what they are trying to accomplish.
Key Differences: IT Networks vs. OT Networks
While Ethernet has become a common Link Layer across both domains, significant differences remain between enterprise IT environments and factory-floor OT environments.
| OT Network | IT Network | |
|---|---|---|
| Terminology |
Devices that make connections are called clients. Devices that receive connections are called servers. |
Devices that provide resources to host devices are called servers. Devices that make connections to servers are called clients. |
| Operating Philosophy | The network is a single-purpose component essential to machine operation and production. Assets outnumber human workers. | The network is a shared utility providing access to resources. Workers generally equal the number of network assets. |
| Traffic Types | Regular, dedicated traffic using specialized industrial protocols with consistent bandwidth. Traffic is localized and can be whitelisted. | Variable, semi-random traffic routed through central enterprise gateways to external internet servers. |
| Performance | Designed for low jitter and deterministic, real-time behavior (down to the millisecond). Systems operate 24/7 without interruption. | Designed for multi-second response times. Traffic experiences peak loads during working hours and low demand off-hours. |
| Monitoring and Troubleshooting | Vendor-specific tools tailored to specific machinery. Few general-purpose troubleshooting tools exist. | Standardized, widely understood network management and software monitoring tools. |
| Network Goals | Safety, reliability, quality and security | Protection of corporate assets, availability, reliability and performance |
| Device Addressing | Devices use fixed IP addresses embedded directly in client memory with zero expected address changes. | Dynamic addressing via DHCP servers. Devices connect using hostnames resolved by DNS. |
| Application Layer Protocols | Highly specific industrial protocols, many of which do not rely on standard Layer 2 Ethernet communications. Examples include Common Industrial Protocol (CIP), Modbus TCP and OPC UA. | Wide range of standard application layer protocols for enterprise data exchange. Examples include HTTP/HTTPS, SMTP/IMAP, DHCP/DNS and FTP/FTPS. |
| Flexibility | Static and locked down. Manufacturers prefer zero changes to devices or software once a machine network functions properly. | Dynamic. Constantly adjusted, upgraded and reconfigured to support evolving applications and users. |
| Failure Modes | Most device failures are generally catastrophic, stopping manufacturing production. Critical manufacturing systems add redundant devices to continue processing after some device failures. | Device failures are generally not catastrophic. Critical routers and switches are installed with redundant capabilities. |
| Risks | Failure can result in bodily injury or environmental disasters. | Failure can impede business operation for short periods of time causing financial damage to the corporation. |
| Device Types | Unique, specialized and sometimes proprietary with long lifetimes. | Standardized on a few operating systems and virtual environments with short lifetimes. |
The Future of IT and OT Integration
It is clear that the age-old lines separating OT and IT are blurred. Ethernet is now the dominant Link Layer in factory floor systems, and digital transformation initiatives are forcing many organizations to either combine their IT and OT teams or at least force them to work together. Differing objectives, standards, tools, missions and the rest will probably continue to make that process difficult.
