What you'll learn from this article:
- How process automation has been driven by collaboration and innovation, from the ISS to Dow Chemical’s MOD systems
- How O-PAS is paving the way for the next era of open automation
As difficult as it is to develop code for today’s digital and virtual process controls, I’m told it was even harder to program and maintain the dedicated PLCs and DCSs of the past. Either way, it’s typically too costly and time-consuming for even the largest end-users to roll their own process controls, so they relied on suppliers for decades. Even the two exceptions I know about tend to prove this rule.
For instance, the last time I covered an end-user creating their own control system from scratch was in 1999, shortly after U.S. and Russian astronauts started building the International Space Station (ISS). Its automation and controls were jointly developed by the station’s international partners, who worked with Boeing to launch about 1 million lines of flight code into orbit on 386 processors in radiation-shielded, multiplexer demultiplexer (MDM) computers. Though they’ve likely been upgraded or replaced in the intervening years, these PCs could be commanded onboard ISS or from the ground. The MDMs ran homegrown code from NASA and Honeywell, managed the station’s U.S. and Japanese components, communicated with two other MDMs handling Russia’s devices, and used a common 15-53 bus network to transfer data among ISS’s computers or even help move the station itself.
Get your subscription to Control's tri-weekly newsletter.
Only slightly more recently, Control ran a multi-part series in 2006 on the Manufacturing Operating Discipline (MOD) process control systems developed organically by Dow Chemical starting in the late 1960s and early 1970s, when few if any other proprietary DCSs were available. Based on its disciplined, continuous improvement philosophy, Dow sought to automate individual loops and manual operations with standardized and reusable controls, which grew from differential equations performed on analog computers to simulate control strategies. Despite hardware constraints, Dow worked with Taylor Instrument Co. (now part of ABB) to build MOD 1, and it was installed on a batch distillation column in 1969.
“Hopefully, O-PAS’s interoperability, decoupling and flexibility will let it upgrade more quickly and easily, and escape the fate of MOD 6 and other short-lived inventions.”
Greater capacity and other improvements were added to subsequent versions, culminating with MOD 5 in 1980. It was deployed globally to 1,500 application starting in the 1980s. MOD 5 saved 45% on control strategy development time, 50% on support costs, and automated Dow’s process startups and batch processes. The company tried to develop MOD 6 starting in 1987, but determined by early 2000 that it was too costly to build its own application-specific integrated circuits (ASIC) and related software. Eventually, Dow got an assist from ABB, and over the years it largely transitioned to its System 800xA DCS.
Will O-PAS succeed where MOD 6 failed?
In the past decade, the quest for less-constrained controls was taken up by the plug-and-play Open Process Automation Standard (O-PAS) and its supporters, who are seeking interoperable process controls, mainly by standardizing the network links between devices, and decoupling software from hardware. This issue’s “O-PAS on the job” cover story reports that ExxonMobil’s first, commercial, O-PAS application at its resin-finishing plant in Baton Rouge, La., is running normally and generating revenue.
And that’s what worries me. Even though this is great news, what happens as soon as an against-all-odds victory is won? One bow and it’s over. Hopefully, O-PAS’s interoperability, decoupling and flexibility will allow its applications to upgrade more quickly and easily, and escape the fate of MOD 6 and other short-lived inventions. Maybe it would help to give up some of our rigid, go-it-alone arrogance? It’s why innovators assume only their solution will be the last word, isn't it?
