Subscribe and listen to AW’s podcast!
Subscribe and listen to the Automation World Gets Your Questions Answered podcast!
Listen Here

Collaboration Improves Efficiency

Standards and new strategies simplify ERP-to-manufacturing links.

Plant operators can access information and relay data to the front office. Source: GE Fanuc
Plant operators can access information and relay data to the front office. Source: GE Fanuc

Efficiency is the watchword in today’s manufacturing operations. A growing number of companies feel that increased collaboration between enterprise resource planning (ERP) systems and the production floor is the way to efficiently boost production.

As more companies look for ways to share information among these programs, there’s a push to transfer data automatically and seamlessly between the groups involved in management and manufacturing. The automotive industry is among those leading the charge in this emerging trend.

“We’re really concerned with eliminating waste and removing redundancy. If we can avoid transferring production counts and other things manually, we can run with fewer resources,” says Robert Barker, manufacturing monitoring program manager at Ford Motor Co., in Dearborn, Mich.

Ford has been working to link the plant floor to various information technology (IT) departments for more than a decade. But for most companies, this type of collaboration has been possible only in recent years, as a variety of software tools simplify communications. After some false starts, the market is starting to grow.

“We definitely see a lot more interest now than a couple years ago,” says Kevin Bernier, director of plant intelligence solutions at GE Fanuc Automation, in Charlottesville, Va.

Negative perceptions

That’s a welcome change for those who want to improve efficiency by implementing ERP, manufacturing execution system (MES) software and other information sharing techniques. Many early efforts tried to impose rigid ERP structures on production facilities that were moving to lean and/or agile manufacturing models. They forced harried production personnel to manually enter data in ways and terminologies that were often foreign to them, leading to some less-than-flattering descriptions for ERP programs.

“For many manufacturers, ERP has become synonymous with failure,” says Robert Lendvai, vice president of marketing at Activplant Corp., of London, Ontario, Canada.

There are a number of reasons that both vendors and users say this negative view is changing, making it far simpler for front office personnel and their ERP systems to collaborate with manufacturing personnel and programs encompassed within MES. Software and service providers stung by negative comments have made major strides to provide more timely information, while standards have improved communications and cut down on the amount of data operators have to input.

Standards have gained acceptance in commercial packages, and even companies such as Ford that use proprietary systems are moving toward standardization. Barker notes that it’s become easier for groups to talk, based on technology improvements and the emergence of standards such as eXtensible Markup Language (XML). That’s true even though Ford began developing its own ERP and manufacturing resource planning (MRP) software more than a decade ago and has stuck with it. “The technology has matured, which allows for easier interfacing, whether you’re using open standards like XML or file type data exchanges,” notes Barker.

This standardization has eliminated one of the key problems that perplexed early ERP proponents. “When you have connectivity, it maximizes what you can do on the production floor to drive down material and equipment costs, enabling the ERP system to get information without forcing people to enter data,” Bernier says.

Another standard seeing more acceptance is ISA S95, an Instrumentation, Systems and Automation Society standard that specifies the data model and best practices for the exchange of information between manufacturing enterprise-level applications and plant-level manufacturing applications.

The trend to openness now extends to open source software. A growing number of vendors are beginning to employ Eclipse, an open source platform for the software used to build development tools. ERP software provider SAP AG, headquartered in Walldorf, Germany, is on the board of directors, and a number of other companies in factory automation are members of the Eclipse Foundation, which took over future oversight and development from founder IBM Corp. early in 2004.

The Foundation has created 33 projects to add new features and functions. One of them, the Eclipse Communications Framework, could help companies move data among team members. “We intend to let people create groups of interest, using ECF to communicate securely between teams. ECF lets people set up communication structures without licensing anything,” says Scott Lewis, project lead for the Eclipse Communications Framework.

Team players

As it becomes easier to move data around, companies are realizing they have to worry about overloading workers. “You have to make sure the right people are getting the right things. Most people don’t want all the data, just what’s important to them,” says Maria Piazza, automation solution director at GE Fanuc.

Determining who gets what information is only one of the considerations that begin with picking the personnel who will make up the team that implements an enterprise-wide solution. Building a cooperative team and getting everyone on the same page is a key aspect for implementation.

“There are many stakeholders involved,” GE Fanuc’s Bernier says. Along with the IT department and production management team are plant managers and vice presidents, as well as process improvement gurus and quality improvement specialists, he adds.

At Integram-St. Louis Seating, which makes seats for DaimlerChrysler mini vans, creating this team was among the most important steps. At times, these teams do something that doesn’t seem to fit the team concept, such as giving one member control of some information. “It takes a lot of coordination. You have to get everyone together early and meet several times. We’ve moved to a single source of ownership of data,” says Gary Fischer, controls engineer at Integram.

That single source ownership is used by many companies. They feel it reduces the chances that erroneous data will create problems. “When there are two sources of information, there’s a chance things won’t line up,” says Ford’s Barker.

Integram is linking its warehousing and material tracking programs to the ERP software, monitoring performance of workstations and issuing replenishment orders when a workstation operator scans material that will be used during that shift. Managers at various levels can see whether a workstation is being underused or is falling behind, and whether an operator is meeting expectations.

Going forward, the company plans to change the way seats are tracked, tracing them by serial number instead of by time and date. That way, if there’s a problem with a seat’s positioning motor, recalls could be more effective.

“We want to be able to go into the system, see which seats have the motors with the problem, then provide Chrysler with the VINs (vehicle identification numbers) of the vans that have those seats,” Fischer says. Doing that will require a lot of input from third-party companies that understand the many software interfaces involved in complex communications among many groups within different companies.

Extending the team concept outside the corporation is something that’s being widely examined by a growing number of companies. Responding quickly to customer requests is gaining importance as product lifetimes shrink and vendors try to offer broader lines and differentiate their products with optional features and functions.

“In automotive, a lot of Tier 1 companies want to link into what their customer uses,” says Piazza. That helps them become more agile, delivering smaller quantities in a more timely fashion, she adds.

Taking small bites

Another key to success is to break the enterprise into smaller pieces that are easier to manage, according to many vendors and users. Many of the early programs failed because they attempted to tackle an entire operation at once. While that’s the ultimate goal, it’s often simpler to devour an elephant by taking smaller bites.

“Companies should deliver visible payback immediately, demonstrating tangible results that set the stage for additional phases,” says Matt Bauer, director of business development and marketing at Rockwell Automation Inc., in Milwaukee.

Once a few of these smaller programs have been implemented successfully, teams can move forward. One of the first steps is to create a universal factory data model that establishes a link between the plant and the ERP system.

“With a data model in place, a bi-directional integration engine can be created that establishes a real-time connection between the plant-floor and the business system. The universal factory data model automatically collects from the entire plant, thus ensuring data accuracy, immediacy and completeness,” says ActivPlant’s Lendvai.

Throughout each of the steps, team members must be sure to know precisely what’s in the software they’re getting. While this might seem obvious, it’s not always something that gets the amount of effort that’s necessary.

“Manufacturing users should be cautious about assuming that key MES functions are incorporated within their ERP systems. For some, finding out the hard way (during implementation) results in elongated deployment cycles, higher integration costs and negative impacts on productivity—in short, a less than optimal solution,” says Bauer.

For more information, search keywords “ERP” and “MES” at www.automationworld.com

See sidebar to this article: ERP-to-MES Link Provides Racer’s Edge

Discover New Content
Access Automation World's free educational content library!
Unlock Learning Here
Discover New Content
Test Your Machine Learning Smarts
Take Automation World's machine learning quiz to prove your knowledge!
Take Quiz
Test Your Machine Learning Smarts