Hershey Co. consumes big data to become a snacking powerhouse
The West Hershey Chocolate Factory can produce 70 million Hershey's Kisses a day. The PI System monitors the process.
The industrial chocolate market might be the sweetest sector of the U.S. economy but like others, it's facing a juggernaut of new business challenges.
The Hershey Company, which holds the largest confectionery market share in the U.S, knows them all. With over 80 brands to satisfy a budding sweet tooth and a Fairtrade savvy connoisseur, Hershey's is dealing with shorter planning cycles, increasingly leaner manufacturing processes, more regulations, and growing demands for a wider product portfolio.
According to a recent report, American consumers are now asking for single-origin chocolates and value-added, premium products at a lower price range. To keep up with new market demands, Hershey's has turned to Big Data and analytics and sought out innovative ways to cut costs and reduce inefficiencies. “Manufacturing is again becoming the core where technology will drive new value streams,” said Russ Gregg, Hershey's Business Transformation Manager for Supply Chain Solutions, during his recent presentation at the OSIsoft Regional Conference in New York.
The PI System at the Sweetest Place on Earth
For Hershey's, many of the answers to today's manufacturing challenges have come in the form of “PI” - the OSIsoft PI System, that is. The company adopted the PI System as its operational data infrastructure more than fifteen years ago at the West Hershey Plant where chocolate lines can make a staggering 2.9 million Hershey's Kisses an hour. The PI System have proven itself to be an effective bridge between IT and OT, delivering not only visibility into real-time production line processes, but proven business insights.
According to Gregg, Hershey's has been able to “leverage real-time data from the machines for its IT applications” for performance analytics. Using OSIsoft's Asset Framework (AF) to standardize data into a consistent hierarchy, Hershey's engineers replicated plant operations into a digital twin of its physical assets. Real-time data from the PI System is fed to AIM, the company's manufacturing intelligence layer that monitors the sweets-making equipment. Before long, Hershey rolled out the PI System to other facilities, including the company's new plant in Malaysia.
“By building that data structure in PI AF, we can eliminate a lot of the complexity from the IT applications perspective,” noted Gregg.
Recently, Wei Chong Lai, an IS Manufacturing Analyst at the Malaysian plant, single-handedly solved a production challenge using the PI System. The Business Unit Leader wanted to track and account for the transfers of cocoa bean paste from the conches that give chocolate its recognizable, mouthwatering taste to the plant's tank farm. The PI System helped to thread the needle. Wei easily imported the DCS control system screens with information about tank IDs, valves positions and pump run states into PI Vision, the OSIsoft's web-based visualization application. Suddenly, his team could see the transfer of various pastes through the system and know the status of each pump, valve, tank, and conche.
“With very little work, you can put the context around [your operations], add a little logic around it, and instantly have insight,” said Gregg. “That's the power of the PI System”.
Today, the PI System is used in five of Hershey's factories, and IT-OT collaboration is no longer optional. The PI System continues to help the company to expand its data-driven culture.
For more about Hershey's and the PI System, watch the full presentation here.