BrooksDay1

Chevron’s Recurring Dream of Technology

April 8, 2008
Using Mature Service-Oriented Architectures Increases Level of Manufacturing Agility.

Mike Brooks has a dream. Chevron’s technology architect presented his vision for the next generation of IT systems aligned closely with business processes and based on shared repeatable work processes in the production management domain, at the 8th Annual Yokogawa Technology Fair & Users Conference in Houston. And he’s looking for vendors that can help him to bring his plan to fruition.

By combining information integration with work process integration, Brooks hopes to create a ubiquitous framework for all applications, where IT functions support the business, and where data access meets the needs of the work or process.

“We now have an opportunity to remove work processes from the IT underpinnings,” says Chevron’s Mike Brooks about the use of maturing services-oriented architectures to facilitate an increased level of manufacturing agility.
“If you really want to get this done in this production management space, you need a framework to support the IT process and a place to build the work processes,” said Brooks. “It has to be a flexible IT architecture. We need an event system with triggers. At the end of this, I want a work process. It looks like a football playbook. As things change, we can put it together in an institutionalized way. We’ve got data modeling, but data warehousing has to be built into the system.”

Institutionalizing custom work processes, to not only lift performance, but also to capture the knowledge of retiring employees, is one of the cornerstones of Brooks’ vision. “How do we get the knowledge from the guy on the third shift who always seems to know what to do?” he asked.

Turnover destroys knowledge and processes, he said. “In the production management space, we have process engineers, and each one of them still does things a different way,” explained Brooks. “When a new guy comes in, he spends two years developing new spreadsheets because the ones his predecessor created don’t make sense to him. Then a new guy comes in, and he does the same thing.”

Brooks advocated collaborative work processes enabled by, but separated from IT underpinnings; secured data within primary systems, but easy, seamless data access and sharing for secondary users; and isolated implementation and vendor packages.

“We now have an opportunity to remove the work process from the IT underpinnings,” he explained. “If I’m in any manufacturing plant, there ought to be one place I can go to find anything I want, whether it’s real-time data or start-up and shutdown procedures.”

The answer to the implementation is in reusable building blocks via services-oriented architecture, said Brooks. “We’re not making shampoo. When we screw up, we get more than bubbles everywhere. Services-oriented architecture is part of the solution, and not part of the problem. Reusable services enable nimble IT architecture.”

To create these building blocks, Brooks sees standards coming to the fore. “We like to use business standards because in the long run we will get lower cost of ownership, interoperability and agility,” he explained. “We need interoperable standards. That’s the end game. That’s the implementation standard. In the open O&M initiative, we make sure the bodies work together. What I’ve been encouraging is for all other companies to get behind this and use this—combining maintenance systems with operations.”

Through open application integration, open maintenance management, open condition management and open reliability management, Brooks sees an opportunity to create an architecture that protects primary users’ data and still allows viewing for secondary users.

When it’s all done, the proposal looks like each system is protected and ensured with one pipe sharing the data to the central repository, which stores data and looks after workflow and can generate query and analysis. The data then can be viewed via secondary user access.

“Our preference is to find outside vendors who understand this problem,” said Brooks. “But it’s got to be more than just a bag of services and workflow. I want a framework that understands how to do this in a repeatable way, just like I do with my DCS.”