Monday, November 16, 2015

The PLC is Dead; Long Live the PAC

By Patrick Weber, MESA International Technical and Education Committee member and MESA-Recognized Practitioner

Will your executive board be surprised at the cost of replacing your automation controllers?

Obsolete controls – every industrial organization deals with them. Whether it’s manufacturing, mining, oil & gas, water/waste water, or building automation, electronic controls don’t last forever and eventually need replaced. I suspect many companies are facing this issue now, particularly as it concerns their programmable logic controllers (PLCs). IEC 61131 (the international standard for PLC programming) was first introduced in 1993, and standards-based devices soon followed. Many of those devices have been in operation for two to three decades, which means we are now in the midst of first-generation controller replacement. This is sparking numerous conversations on social media, such as this one on LinkedIn’s Automation group forum: “What about automatically converting PLC software from one system to another”.
It’s clear we are reaching the end of the era where machines were isolated cells of production; modern automation is expected to be far more integrated into the business processes in ways first generation controllers were not capable – leading to “programmable automation controllers” (PAC). They must provide data to facilitate maintenance. They must interact with enterprise resource planning, manufacturing operations management, and product lifecycle management systems. They may be required to collaborate with other machines, materials management systems, robotic systems, and quality management systems as well. Automation controllers are no longer limited to managing machine behavior – they have become core components of manufacturing intelligence.
Obsolete controls are an example of what’s known as “technical debt” in the information technology world becoming exposed in the operational technology field. When a device (a PLC, for instance) is first commissioned, it begins to accrue a debt that will eventually need to be paid. The environment around the device changes, changing the problem set for which the device was the intended solution. Sometimes an organization will elect to re-engineer the process to adapt to changing conditions, but the more typical response is to “tweak” the system and delay the eventual expense and effort required, increasing the debt payment. Just as with financial debt, this isn’t necessarily a bad thing – but unmanaged debt usually leads to undesirable consequences.
In operational technology, technical debt has broader reach than is first apparent. Beyond the cost of the controller, the software to program/maintain it (plus the training to keep personnel current), and reprogramming/commissioning costs, there is the surrounding infrastructure to consider, from I/O wiring to networking to support servers and services (OPC, revision control, etc.). Modern controls are typically accessible via a corporate network, even when isolated using technology such as V-LAN or DMZ; security has become a greater concern. There are also newer industrial standards which should be included as part of the debt analysis; it isn’t just the controls hardware which has become obsolete. There are modular programming standards (ISA-88/IEC 61512), integration standards (ISA-95/IEC 62264), and automation system security (ISA-99/IEC 62443) which should be incorporated as part of a complete controls update program. The standards continue to evolve as well; the ISO Automation Systems and Integration Technical Committee (TC-184) is working to adopt the ISA/IEC standards as well as adding standards around robotics and industrial data management.
Obsolete controls programs should cause industrial organizations to pause and reflect on their operational strategies; what do they want their production processes to look like five to ten years in the future? What will their competitor’s processes be like then? Such questions will have an impact on decisions such as “do we spend the time to re-write the PLC code for a new processor, or do we just do the minimum work required to fit the new device into our existing processes?” If those questions are not asked, automation managers will choose the most direct route to keeping production processes operational. They will elect to increase the technical debt.

No comments: