You are here
This simulation of a droplet of liquid falling into a pool of liquid was modeled using Los Alamos National Laboratory's Computational Fluid Dynamics Library (CFDLib), which was also used by Procter and Gamble to simulate a manufacturing process. The computer code is now available to help American industries become more competitive. | Courtesy of Los Alamos National Laboratory
Procter & Gamble (P&G)’s laundry detergent packaging line is immense — from the mixing vats to the filling machines, which then inject empty bottles with the detergent. Conveyor belts whisk the full bottles to capping machines and then to labelers, cartoners, and the machines that put the boxes onto pallets. Just one component breaking or a product jam along the complex system can stop an entire line. Failure means substantial profit loss: packaging line stoppages cost P&G more than $1 billion each year.
That was before P&G partnered with the Energy Department's Los Alamos National Laboratory (LANL) in the 1990s. LANL scientists helped P&G engineers develop simulations to improve the reliability of P&G's complex production lines. P&G's 150 facilities worldwide saw a 44 percent increase in plant productivity and 30 percent increase in equipment reliability since they started using the software.
Large-scale implementation of the technology helped save P&G $1 billion in manufacturing costs, according to Procter & Gamble. These cost-saving benefits are applicable towards production lines across the manufacturing sector, which was responsible for $1.7 trillion of the US GDP in 2010, according to the Bureau of Economic Analysis.
But like many early technological advancements, the commercial application wasn't initially clear. LANL researchers had developed advanced statistical models to quantify uncertainties in complex systems, but they didn’t have enough real-world data to know whether they were correct. At the same time, P&G had a lot of operational data from its complex production lines, but no model to use them effectively.
The pairing of the lab and corporations' data led to the creation of simulation software called Reliability Technology in 1993. With the software, engineers can configure both the machines and their maintenance schedules based on reliability. In addition, engineers could foresee and possibly avoids product jams, intervals of a component breakage or variations in a machine speeds. In other cases, engineers could triage the production line.
Two years later, P&G field-tested the program with success. The statistical model revolutionized the sliver of systems engineering called reliability engineering. Through the real-world data sets, LANL verified their models. The lab went on to form similar partnerships to support 15 other projects to develop LANL technologies.
In 2001, eight years after the initial pairing with the lab, P&G commercialized and licensed the product as part of a manufacturing operations toolkit. Touted as a way to keep the machines running as fast as they could for as long as possible, the software also improved machine reliability, lowered the rate of waste and created safer work environments.
The LANL partnership with P&G leveraged both industry and research goals to not only more efficiently produce products, but strengthen the nation's economic competitiveness.