Skip to main content
Industrial Internet

Taking Stock: This Software Is Taking The Pain Out Of Inventory

Dorothy Pomerantz
Bruce Watson
October 16, 2017
In the perfect business world, there would be no red ink, no broken machinery and no needless meetings. Many factory managers would also add this to the list: no inventory gathering dust on the shelf.
But the real world is still pretty messy. GE’s 500-plus factories and manufacturing sites, for example, hold billions of dollars’ worth of inventory. Those compressors, propellers, fuses, resistors and other parts take up space in warehouses that must be kept staffed, lit and maintained. All that money could be put to work more productively somewhere else.

One way to measure inventory efficiency involves counting the number of “turns,” the time it takes for a stock to completely turn over. At GE, each turn represents $3 billion tied up in inventory. For Francisco Montecinos, that was a rich target.

Montecinos is vice president of supply chain solutions at GE Digital. Last year he wondered if the company could use an application based on Predix, the software platform GE developed for the industrial internet, to help its businesses reduce turns by three-quarters, a $2.25 billion opportunity. He partnered with supply chain leaders across GE to better understand how to make the process more efficient.

Inventory maintenance is a delicate balance: You need to have parts on hand, but you shouldn’t let them sit around too long. To manage this, factories use planning and forecasting tools. Businesses review and anticipate what they will need for the next cycle and work with suppliers accordingly.

 width= Inventory maintenance is a delicate balance. Top and above images credit: Getty Images.

But as the saying goes, everyone has a plan until they get punched in the mouth. A customer calls and says they need a product sooner or later than expected, or the supplier calls with a change of plans, and suddenly parts that you need are not there, or parts are stacking up on shelves. “Everyone we talked to said they needed a better way to manage the material flow,” Montecinos says.

About 12 months ago, Montecinos and a team of supply chain and digital experts set out to improve flow based on a three-step process: First, take information coming from software that factories already use, like enterprise resource planning tools, and stitch it together. Then, use the power of Predix-based analytics to analyze all of that data and offer real-time recommendations. Lastly, present those recommendations in a way that is tailored to each step along the supply chain.

Too often, for example, changes to the schedule do not reach all the right players within the factory on time. To address this and other problems, the team developed an internal application with multiple modules, each targeting a specific “persona” such as a buyer or a plant manager. Each module covers a different area of the supply chain. The modules use data pulled from existing factory software and analyze it with Predix analytics. A planner who schedules production, for example, can use it to make sure production meets customers’ needs or highlight overdue orders. A buyer gets recommendations based on a different set of modules than a plant manager or a scheduler. But all those recommendations are using information that is being gathered across the modules.

“Based on all of the supply chain experience we have at GE, we are able to capitalize on the deep domain expertise across the company, and combine that with the power of digital to provide actionable insights,” Montecinos says.

When the team developed the software, they first tested it at two GE factories. It allowed workers and managers make better decisions and led to a 10 percent decrease in inventory. It now connects more than 80 plants across GE. Raj Thakkar, GE’s vice president of supply chain integration and a champion for the project, says that in a "dynamic environment" like the supply chain "things change all the time. The key is to have a digitized standard of work and seeing actionable signals apart from noise," he says. "Now we have a better, smarter and simpler solution that enables and advises the materials team to make the right decision at the right time.”

Up next: Montecinos' team is working on modules that dig deeper into software automation, explore machine learning, and offer more advanced recommendations and actions.

Says Montecinos: “Streamlining inventory is a key part of improving cash flow.”