Procter & Gamble Highlights its Digitization Strategy

By Jamie J. Gooch

P&G’s Laura Michalske

You might think one of the largest consumer product companies in the world, with household names such as Pampers, Crest and Tide would not have much in common with heavy industry when it comes to manufacturing and simulation. Laura Michalske would beg to differ. Michalske, who has the impressive title of engineering section head for global health & well being modeling & simulation with Procter & Gamble, provided insights into the company’s manufacturing, simulation and product lifecycle management (PLM) processes during the May 25 keynote at the NAFEMS World Congress in Boston.

More than 90% of U.S. households have a P&G brand in their home. The 174-year-old company serves about 4.2 billion people in more than 180 countries. So what its products may lack in overall complexity compared to, say, all the working parts of a jet or an automobile, the company makes up for in volume.

“At P&G we make over 1 million data updates every day to product and package changes,” said Michalske, who is also chairperson for the NAFEMS Simulation Data Management Working Group. “Our challenge is to leverage our scale while avoiding complexity due to size.”

To drive home that point, she compared the number of submarines a company might have to make before attaining $1 billion in sales: 1, or jet engines: 1,000 to the number of product sales P&G has to make to earn $1 billion in sales: 7 billion.

Aerodynamic Pringles
That’s a lot of products. According to the P&G website, the company has 23 brands that generate more than $1 billion in annual sales. And of course, the faster it can manufacture products, the faster it can make those sales. P&G attains that speed with automation via modeling of production lines and the products on the line.

For example, the company makes a billion diapers in just a few days and a billion Pringles in just a few hours. Those speeds create challenges.

“It’s an aerodynamic problem,” Michalske said, referring to the speedy Pringles production line. “That chip wants tofly off the conveying line. We actually had to model the aerodynamics of a Pringle.”

As recently as 10 years ago, when the company wanted to introduce a new product or change the packaging of an existing product—say a bottle of detergent—it created a physical prototype of the bottles and the “racetrack,” as P&G calls the conveyor line that speeds products through manufacturing. If there was a problem with the test run, such as the bottles tipping over on the line, the packaging and/or racetrack would need to be redesigned and reproduced for more testing. They basically had to start from square one.

“Today, we do it virtually,” Michalske said. “We do the bottle design virtually, we do the racetrack virtually. We do that in very little time compared to how we used to.”

The same holds true for mixing liquids products. The company uses computational fluid dynamics (CFD) to ensure solutions will mix well. It uses finite element analysis (FEA) to ensure its packages can handle being stacked atop one another on pallets or dropped by consumers.

Realizing the Promise of PLM
As you might imagine, the amount of data coming from those simulations was massive. The company soon realized it was not using the data effectively to reduce its costs.

“A significant amount of organizational energy is spent creating and managing knowledge and data,” Michalske said. “Modeling a simulationis creating larger data sets and results to manage. Global organizations such as ours need to have the right data at the right moment to innovate products and production systems.”

The company began using PLM in the 1990s, but wasn’t seeing the benefits it expected. It found the cost of maintaining PLM to be problematic.

In 2007, the company staged a PLM intervention to develop a new plan to deliver the full PLM promise. The company plans to digitize from “atoms to enterprise” by 2014 with its strategic partners, Dassault Systemes and Siemens.

“We have connected M&S (modeling and simulation) results to business decisions made, such as having a digital record of which analysis drove which decision for IP (intellectual property) and to minimize rework,” Michalske said. “We extended the reach of M&S—more people performing more analysis—and have improved M&S analyst productivity overall via workflow templates and digital collaboration between analysts.”

She said the company realized it needed to integrate SLM (simulation lifecycle management) into the PLM workflow if they wanted more analysts to use it.

“The combination of M&S and PLM and SLM (simulation lifecycle management) at P&G is how we’re changing our work,” she concluded.