The Hershey Company turned to Internet of Things (IoT) and machine learning to gain production efficiency. Before implementing the IoT solution in Twizzlers factory line, the company was facing significant challenges in regulating temperature of holding tanks during the manufacturing process. To comply with minimum legal sizing guidelines, the company had to make its candies slightly bigger.
“I can’t give you 1.99 ounces, so I give you 2.19 ounces”, said George Lenhart, senior manager Workspace solutions for Hershey’s during his IoT World presentation. This guesswork approach for sizing the Twizzlers resulted in obvious production inefficiencies.
The company started their IoT improvement project by identifying existing sensors on the manufacturing plant. It then added new IoT sensors to monitor the temperature of the hot Twizzlers in each holding tank. 22 temperature sensors from the cooking vat were used to assess the temperature every second of every minute while the licorice was in the holding tank. Additionally, the company mounted intelligent sensors on the factory line extruders.
This allowed the company’s data team to collect 60M data points. Temperature, torque, and pressure data was tracked using the intelligent sensors. The data was then transferred to Azure cloud where precooked algorithms inside of Azure machine learning were used to build, deploy, and share predictive analytics about the machines (holding tanks, extruders, and chocolate-making Conch). Using Microsoft’s Power BI, the data gathered from sensors was streamed to a dashboard to monitor extruder performance in real time.
Insights gleaned from Azure’s machine learning algorithms were used to reduce variability in the sizing of Twizzlers and improve extruder performance. The company was further able to save hundreds of thousands of dollars in decreased waste and create a leaner operation.
