What Is OPA? Part 3

Show notes

This episode continues The OT Crowd’s What Is OPA? series with a focused look at architecture. Jarie and David break down why open process automation is built around standards, loose coupling, and separation of responsibilities so industrial systems can evolve without constant rip-and-replace upgrades.

David explains the core architecture by comparing traditional PLC and DCS setups to modern computer systems, then shows how OPA separates control, communications, advanced computing, and orchestration to improve flexibility, resilience, and lifecycle management.

Watch on YouTube. Listen on Spotify or Apple Podcasts.

Key topics

• David explains that a PLC or DCS is not a magical black box, but a specialized computer with an operating system, applications, network ports, and APIs.
• The discussion centers on why industrial systems need to be treated differently from ordinary IT systems because they control machines with energy, momentum, and potentially hazardous materials.
• David highlights the lifecycle mismatch in automation systems: field wiring can last 80 to 100 years, I/O may last decades, software may need replacement in 10 to 15 years, and computers often age out even faster.
• A major theme is the problem of tightly coupled systems, where hardware, software, I/O, and form factor are all bound together, forcing full system replacement when one part changes.
• The episode contrasts that with OPA’s goal of loose coupling, abstraction, and standardized interfaces so components can evolve independently.
• David introduces the distributed control node, or DCN, as the OPA term for the controller layer that performs control functions without being tied to traditional PLC or DCS assumptions.
• The communications layer is described as the OCF, or Open Process Automation Communications Framework, with OPC UA chosen as the protocol standard for data, cybersecurity, client-server, publish-subscribe, and time-sensitive networking needs.
• The advanced computing platform, or ACP, is presented as the server-side environment for engineering tools, historians, optimization, analytics, and other non-control applications.
• Orchestration is a major highlight, defined here as automation of the tasks needed to deploy, configure, secure, and join systems to the network, not just simple scripting.
• Jarie and David emphasize the operational payoff: faster deployment, better redundancy, less single-point-of-failure risk, and lower total cost of ownership.

Timestamps

• 00:00 — Part three of the OPA architecture series
• 00:54 — Why OPA starts by treating controllers as computers
• 02:16 — What is inside a PLC or DCS cabinet
• 03:11 — Why OT systems are not the same as office IT systems
• 05:07 — The automation lifecycle mismatch: wires, I/O, software, and computers
• 06:34 — How today’s cutting-edge technology becomes tomorrow’s junk
• 07:32 — How tight coupling creates the rip-and-replace problem
• 08:30 — Why loose coupling and abstraction are the real goal
• 09:59 — The distributed control node, or DCN
• 10:55 — The OCF and why OPC UA was selected
• 11:53 — Control, engineering, historians, and the advanced computing platform
• 14:18 — Orchestration as automation for deployment and security setup
• 15:54 — Bringing up a DCN in 18 minutes
• 17:49 — Why controller and I/O do not need to know each other directly
• 18:47 — Why separating I/O from compute improves resilience
• 19:47 — Using industrial computers for optimization and local logging
• 21:15 — Protecting control, visibility, and data across upgrades
• 22:12 — Scaling systems without being trapped by vendor limits
• 23:16 — Closing thoughts and wrap-up

Notable quotes

“The PLC and DCS is a computer.”

“Don’t fall in love with your technology because today’s cutting edge stuff is going to be somebody’s junk someday.”

“The system is for the end user. It should grow and evolve and change as the needs change.”

Next: What Is OPA? Part 4.