Making Data Talk

As manufacturers install more intelligent sensing and computing devices around their plants and factories, questions arise about what to do with all the data.

Aw 2650 0909 Tech

Industrial networking, especially Ethernet with Internet Protocol (IP), gives a boost to moving data. Suppliers have stepped up with a variety of database products and historians to store all the data. But the problem remains—how to retrieve and use the data.

Data mining is a concept for retrieving this data, looking at them from various points of view, deciphering patterns and providing visual displays for users. In this way, people can turn the data into actionable intelligence. “Data mining is important for any organization, as end-users can begin to think outside the box and deliver creative answers to today’s business problems. Users become proactive rather than reactive, which results in better informed decision making, improved transparency throughout the business, and overall, a more efficient and effective organization,” says Tristan Ziegler. He is chief executive officer of Visual Mining Inc. (www.visualmining.com), a Rockville, Md., data mining product supplier.

Hidden patterns

Data mining is the process of extracting hidden patterns from data, says Ziegler. And, he continues, performance dashboards offer a way to quickly and easily mine data and scale it to make better business decisions. Users control design and layout, so they don’t need a programmer or information technology (IT) department to implement the solution, and they don’t need IT to customize the dashboards for personal preferences. Dashboards can be effectively used for three-week projects as well as for three-year projects.

Ziegler says an effective performance dashboard enables businesses to:
l    Turn mined data into actionable information
l    Improve the understanding of business data
l    Gain faster time-to-value from technology
l    Develop more effective business operations.

Kevin Scott, Visual Mining’s vice president of engineering, adds, “We’ve been in business for 13 years, but the phrase ‘data mining’ is a new concept. But in general, there is a lot of corporate data that people need to get to quickly—and without a lot of overhead. We don’t want the process of getting data to get in the way of actually getting the data. Some projects, though, are large and could take months with a big budget. Some companies have sophisticated requirements that are hard to implement. The state of the art is to take an application and point it at the corporate data warehouse and tell it to go find something interesting. Then it comes back and says, ‘Sixteen-year-old boys buy blue shirts on Tuesdays.’ ”

Visual Mining provides information in a graphical manner with interfaces to allow users to explore and drill down into details. Customers use the tools in a variety of ways. For example, says Scott, “We went into a prospect site and they said, ‘We’re concerned with competitor A and competitor B. We want to know why we’re losing contracts and why we win contracts.’ So, we pointed our software at their data warehouse and did a study. The interesting thing is that when the information came back, it was competitor C who was eating their lunch.”

Another example involves a silicon chip manufacturer that wanted to analyze defects in raw wafers it was getting. “They use our software to assure the raw materials they are processing are good. The process is too expensive to work on bad silicon,” Scott relates. “Exploiting all the data in your data storage is a trick. Data mining is a way to do it.”

Gary Mintchell, gmintchell@automationworld.com, is Editor in Chief of Automation World.

Visual Mining Inc.
www.visualmining.com

Subscribe to Automation World's RSS Feeds for Columns & Departments

More in Control