How to choose manufacturing software for business intelligence
A comprehensive collection of articles, videos and more, hand-picked by our editors
Embarking on business intelligence (BI) initiatives can be intimidating for a manufacturer -- or any company, really. There are always burning questions, such as: What is the root cause of our shipping delays? How can we reduce our defect level to zero parts per million? BI software often promises to be the silver bullet to solve these challenges.
At the same time, many manufacturing executives view BI as never-ending, complex and expensive. There is some truth to those fears. If you're serious about getting into BI, you should take a methodical approach and create a strategy long before you evaluate one tool. The strategy should incorporate everything from a road map for technology implementation to what questions the new system will answer, experts say.
Most important, however, a BI strategy should take into account the organization's cultural readiness for analytics. Even more so than with other large IT projects, change management issues can tank BI initiatives.
"BI involves a complete culture change," because employees are able to make performance data-based decisions, often for the first time, said Jeff Seabloom, managing director at Alsbridge Inc., an advisory services company with offices in Dallas and London. "You need to have people and teams ready to embrace it." A good starting place, he added, is communicating why the organization needs the analytical capabilities, as well as discussing information they would like to possess and questions they would like to answer.
Recycling company prepped users for BI initiative
Brian Hanley, senior BI architect at Gopher Resource LLC, understood the magnitude of the undertaking from the outset. The Eagan, Minn., recycler of lead waste needed better insight into the performance of its furnace, the heart of its operations. Before implementing a BI tool, though, Hanley needed to make sure the users were comfortable with the concept of analytics.
"We had a user base that had no exposure to analytics. We did not want to put the cart before the horse," said Hanley. A critical first step for driving organizational readiness was the creation of a data store, a relational database against which users could run queries.
From a technical standpoint, the data store is below a true data warehouse -- which Gopher plans to build next year -- giving the company the ability to do advanced analytics using cubes. "We are ready to mature and grow into this. [Employees] have now become accustomed to making decisions based on the performance data," Hanley said.
Hanley's group also worked to identify and remedy user resistance to the BI software before it had a chance to harden. Long accustomed to using Excel, many users did not understand why they couldn't keep using the well-entrenched business tool.
"We identified different user groups and worked with them to understand why they wanted to keep using Excel," said Hanley. His team was able to show that continuing to use the old tool in avoidance of the BI system would create another source of data, undermining the organization's need for better understanding of its production information.
"We immediately tried to correct whatever the real or perceived deficiency was, because that can really undermine confidence in the system," Hanley explained.
Poor company culture steps on BI's toes
Not every manufacturer is so methodical. Mark Cox, director of technical training and advanced systems at Advanced Technology Services Inc., has found in 30 years of consulting that most manufacturing companies are not prepared to change.
Cox recently took on the task of identifying improvement opportunities for a large manufacturing company that had just spent $100,000 on a data collection system for a struggling production cell. While the company was collecting data on the shop floor, employees were reporting overall equipment effectiveness (OEE) levels exceeding 100%. This did not make sense, given the cell was not meeting production or cost targets.
After review of the data from the new system, Cox discovered the true OEE was well below 30%. Unfortunately, no one on the production staff was willing to publish that number, fearing it would make them look bad. Meanwhile, Cox was astounded that management had apparently not noticed the reported OEE levels, a clear indication that something was amiss.
"If I would have known the culture within that company prior to my visit, I would have recommended that they save the $100,000 and spend $100 on pizza for the shop floor employees," Cox said. "At least the company would have gotten some good will out of the investment and the production improvements achieved by the data collection and analysis would have been exactly the same with a lot less cost and hassle."
For manufacturers like Gopher that take the time to prepare employees, benefits are within reach. "We have been able to take a more measured approach, and it will set us up well for success," Hanley said.
Follow SearchManufacturingERP on Twitter @ManufacturingTT.