jonathan-reichert-tyson
jonathan-reichert-tyson
jonathan-reichert-tyson
jonathan-reichert-tyson
jonathan-reichert-tyson

Improved visibility at Tyson Foods exceeds ROI threshold by factor of 5

Jan. 15, 2016

Tyson Foods has realized significant improvements in sausage line performance through improved visibility. Jonathon Reichert of the company's Hillshire Brands division explains how.

Food safety drives many capital investments in food production, and safety helped spur an Internet of Things (IoT) project at the Hillshire Brands division of Tyson Foods, according to Jonathan Reichert, senior engineer-innovation in Hillshire’s corporate engineering group in Downers Grove, Ill. Reichert discussed the company's successful implementation at the Smart Industry 2015 conference.

Archiving production data is important: one of the firm’s sausage-cooking facilities was forced to shut down because managers were unable to validate the temperatures within an oven. Temperature and hold time are critical variables in validating pasteurization of processed foods.

With a data historian and visualization tools from Rockwell Automation in place, engineering could begin leveraging the new infrastructure for other initiatives. The first involved 1 lb. chubs of Jimmy Dean sausage, a billion-dollar brand and Hillshire’s top seller. Lean-to-fat ratios and total weight are key variables that must be monitored, and they are among the 2,500 data points collected in the historian.

Selling a chub that weighs 17 ounces represents product giveaway, but selling a chub weighing 15.9 ounces would invite regulatory wrath. After chubs are clipped, they pass over a checkweigher, where underweight chubs are rejected. Until the IoT system was installed, weight data was never recorded.

With 11 production lines outputting 100 million lbs. of sausage a year, Reichert calculated a 0.1 percent improvement in targeted vs. actual chub weights would justify the plant’s investment. The goal was achieved in three months, and the application of statistical analysis to identify drift in real time has now resulted in a 0.5 percent improvement.

[callToAction]