Creating the perfect wine that blends a flavored bouquet of "ripe apple and pear with hints of butter, toasty oak, and spice" requires all the right conditions: the soil, the climate, the time of harvest. Then there's the crushing of the grapes, the fermentation, and the bottling. If all goes well it adds up to a top-selling super premium chardonnay. A consumer sipping the tasty libation is not likely to grasp the integrated effort required to produce the wine. But Andy Woehl is one person who fully understands what it takes to manufacture a great wine and, more importantly, the competitive need to improve the process. As the plant process engineer for vintner Clos du Bois (Geyserville, CA), Woehl over the last two years has focused on connecting all of the company's wine-making production systems in order to extract critical information from them. He's connected the barrel production system and the automated grape-crushing system using Rockwell Automation control technology and an Ethernet LAN. More recently, he's implemented bottling-line performance analysis software that works with Rockwell's manufacturing execution system (MES) to create sophisticated reports. The automation helps move data that can be used to assess and improve key processes. But that's just a start. "Data alone does not improve bottling efficiency," Woehl says. "It's really how we as a team use the data." Woehl, like many manufacturers on a progressive path, realizes that team effort is the key to a successful data integration project. The technology to tie things together is readily available in the form of communication networks, middleware, application interfaces, and Web services, for example. But technology is just a tool to move data around the plant or even up into the enterprise. The real value of a data integration project comes from people, processes, and continuous improvement practices. "When doing [data integration] projects, it is very important to think more holistically than just technology," says Paul Loftus, managing partner of industrial equipment in North America for Accenture (Chicago). "How well you use the information is pretty important." Integration efforts require a single operating model, clearly defined metrics, and, most importantly, the leadership to instill a cultural commitment to the project. "At the end of the day if you look at the people who are trying to drive shareholder return and value, they are looking hard at how the operations of the company are going to run, how to drive consistency and facilitate a culture of continuous improvement from fact-based information," Loftus says. Change Agent All technology projects come about because a company has recognized the need for change. At Clos du Bois, Woehl's goal was to pull more information out of a bottling line that was consuming a lot of money on a daily basis. When he started at the company five years ago as a lab technician, no data was being captured from the production process. Everything related to production performance was a subjective argument with no clear view of what was really going on. Since implementing the line performance management system, he's been able to capture enough data to show that the average year-round bottling efficiency hovers around 65%. "We'd like that to be at 80% efficiency," Woehl says. That's where new processes will have to factor in. But it's still a work in progress, he adds. Automation hardware and software vendor GE Fanuc (Charlottesville, VA) has been focused on improving processes through data integration and analysis a lot longer than Clos du Bois. But, like the vintner, Fanuc has discovered that the key is people and teamwork. Specifically, according to CIO Liam Durbin, Fanuc has learned that one secret to a successful integration project is to involve key people from all parts of the business, not just IT. The company uses the Six Sigma continuous improvement principles to do this. General Electric, the parent company of GE Fanuc, adopted Six Sigma quality processes in 1995. The Six Sigma philosophy -- which in real terms means that the "occurrence of finding a defect is 3.4 defects per million opportunities" -- has been driven across all GE business units into departments outside of manufacturing and distribution, including finance, IT, and human resources. In order to reinforce the effort, Black Belts, or functional leaders, are trained on business metrics and are responsible for measuring, analyzing, improving, and controlling key processes. "A Black Belt is a change agent," Durbin says. "A person trained and then given a major initiative which revolves around a metric, like lean." Since taking over as CIO in 2004, Durbin has more than doubled the number of Black Belts under his watch to 28. The other important aspect in any GE Fanuc data integration project, Durbin says, is moving through what he calls a toll-gate process. It's a way to make sure that no one is left out of the discussion. Thus if someone's needs are not being addressed, the project will not go forward. A good example of how GE Fanuc operates is illustrated in its recent upgrade to version 7.8 of Siebel Systems Inc.'s customer relationship management software. "Instead of having just the IT people in here for the weekend, we had the entire business team. We weren't going live until it was tested 100 percent," Durbin says. "We had functional owners in here with carefully scripted, exhaustive tests." That inclusive approach, Durbin says, delegates responsibility and mitigates risk. "It goes from being an IT project that you throw over the fence and say, 'Let us know when you're done,' to being a business project with functional ownership, which is what makes sure you get the project over the business line." All Together Now Because integration efforts are typically cross-departmental and cross-functional, having organizational cohesion should be a corporate imperative. It comes down to leadership and a commitment to the operating model, says Accenture's Loftus. "You need to drive some level of consistency across all of your plants [in terms of] operating performance and metrics so that you can compare them." But creating that consistency can be a daunting task these days. Acquisitions, for example, forcibly fuse IT systems while also creating companies composed of multiple manufacturing sites that are often dispersed across different geographical regions. Increasingly, manufacturers are looking to their technology vendors to help them develop and deploy consistent operating models. "We are seeing a huge increase in what customers expect from us as a technology provider," says Claus Abildgren, program manager for production and performance management solutions at Wonderware (Lake Forest, CA). Abildgren maintains that Wonderware is not a systems integrator, but, due to increased pressure from confused customers, the company offers consulting services to help companies deploy technology in the right way and, more importantly, in a standard manner. "They want us to create corporate standards, not just as a single application point of view, but defining metrics and business processes that should be captured in technology and designed as templates for reuse." A consultant can help steer efforts to put in place a standard, consistent operating model. But to really work, such initiatives must be driven by top management. Therein lies a major problem for Clos du Bois' Woehl. The company was recently sold to Fortune Brands Inc. (Lincolnshire, IL), a large marketer of not just spirits and wines but also golf and home and hardware products. Fortune's top priority is brand awareness. Manufacturing, on the other hand, falls to the wayside, particularly when it comes to budgeting. As a result, Woehl says, technology initiatives are often pushed from the bottom up. "In the IT world that means the project can be impossible at times to get completed," he says. That's why Woehl is aggressive about getting the entire team involved and ensuring there is a sense of ownership among the ranks. Line of business managers as well as IT managers must share that sense of ownership, Woehl says. It can be a tough sell, especially when the extraction of data yields greater visibility into the production cycle. "Clos du Bois management saw it as a tool, but production saw it as a possible blame game. We have to change that mindset," he says. Understanding how time is being spent in the manufacturing process is an important element in any data-driven initiative. For example, at Clos du Bois, corporate policy states that the bottling line starts at 7:00 a.m. each day. But the reality is that 7:00 a.m. may actually be the time the workforce is getting situated, which means expectations of production are slightly skewed. Insight into operations shouldn't lead to finger pointing, but rather a baseline for correction. The reality that bottling doesn't actually start until 7:15 a.m. or so "means we either have more procedures in place that we are not giving enough time for every day, or maybe we should just say we start [bottling] at 8:00 a.m.," Woehl says. The situation at Clos du Bois is all too familiar to many manufacturers, some of whom are turning to tools that integrate information about human resources with production data. "The first step is getting visibility into what the workforce is actually doing," says Deborah Torchia, director of product marketing and product management at SmartTime Software Inc. (Framingham, MA). "Then, as you make process improvements, you start to see where the processes are not working." SmartTime is not a typical time and attendance solution. Rather, it can record transactions -- such as when an employee swipes a card at a Web kiosk to set up a workstation. Data about the event is captured in a backend system that maps activity codes and labor levels to a company's general ledger system. Pulling it all together provides one consistent version of the truth. That kind of integration also offers valuable insight into employee activity. For example, overtime is often a symptom of a problem within the process. "Getting that level of data can solve the root cause of the problem, making production lines more efficient and margins better," she explains. Woehl is not depending on a workforce management product since Clos du Bois has only about 100 employees. Twenty of them run the two bottling lines, each of which produces approximately 320 bottles per minute. Instead, he and about five other project leaders meet every Friday to go over the information they've captured for the week. "We decided to collaborate as a group and come up with a week-by-week list of possible ideas [for improvement]. We want them to think, analyze, offer ideas... and to feel like they are being received and applied," Woehl says. Integration, after all, means involvement -- an idea worth raising a glass to.