"Smart manufacturing" is an umbrella term that encompasses a range of developments, including the use of advanced technologies, the internet of things and the application of analytics to make manufacturing more productive and responsive, says Jacobson. It's not accurate to call it a revolution, though. "It's more of an evolution. It’s the next level of cost and productivity gains."
The concept is nevertheless a new one. In the past, Jacobson says, plants have had connected devices, but they were largely internal in nature – an "intranet of things." Manufacturers today are just realizing how to make the best use of data, 70 percent of which is generated during production and never gets outside the plant.
Those techniques are critical in today's manufacturing environment, which is characterized by a substantial amount of localized production in support of more varied product mixes, and an overall increase in complexity. Producers are "trying to do more with less," says Jacobson, as they labor to be more creative with their factory setups.
In some cases, machine learning can be of help, with automation deployed to avoid "process drift," and to draw on customer feedback in order to improve product and process quality. The challenge, says Jacobson, lies in making real-time use of that critical data.
In the age of Big Data, many companies find themselves inundated with information, and the difficulty of ascertaining what's relevant. But factories have always produced large volumes of data, Jacobson says, even if much of that was in the form of "dark data" that went undetected and therefore unmined.
The promise of big data becomes a reality once a factory is connected with the larger supply chain, ensuring tight relationships between product performance and manufacturing, Jacobson says.
