Think about dairy and you likely conjure up familiar grocery staples like milk, cheese and butter. But the proteins found in raw dairy product also serve as a key ingredient for a wide range of health and nutrition products.
Milk Specialties Global is a major supplier of these proteins, with 10 manufacturing facilities located throughout the Midwest and in California. The Minnesota-based company processes raw milk and whey into a broad portfolio of ingredients, including milk proteins, whey proteins and specialty proteins.
These ingredients are then shipped to customers that use them in food and beverage products ranging from coffee creamers and baking goods to processed cheeses and Greek yogurts. They are also a primary ingredient in high-demand, sports-nutrition, weight-loss products, such as protein powders, protein shots and meal-replacement bars. And they are used in a variety of animal-nutrition products, including milk replacers, energy boosters and nutritional supplements.
Milk Specialties has rapidly grown in recent years to keep up with market demand. The company of only about 800 employees is today nearing $1 billion in annual revenue.
To help meet high demand, the company recently made investments in manufacturing intelligence to achieve real-time analytics and insights into its operations to drive improvements in production processes.
Untapped Data, Information Deficiency
Despite the tremendous variations that can occur in the raw-ingredient input streams that the company receives, it still must deliver consistent-quality output streams.
For its human-nutrition products, the raw, dairy-product stream is separated with state-of-the-art membrane technologies– which allows the isolated proteins to remain in their native state. From there, the stream is sent through a dryer and finally to the packaging line.
Identifying and improving inefficient processes in these operations was challenging. The facilities’ production control systems were data silos, with little ability to track or compare data over time. And the reporting capabilities that were in place were outdated and labor intensive.
For example, tracking the clean-in-place (CIP) chemical usage involved workers downloading data from a machine into a Microsoft® Word document and then manually entering that data into a spreadsheet to create a graph. This not only took time, but it also restricted workers’ abilities to view data quickly and within multiple contexts to truly understand CIP chemical usage.
When the company inherited a legacy, historian-software system with an acquired production facility in Visalia, California, Mike Calva, corporate programmer and analyst for Milk Specialties Global, saw an opportunity to do more with production data.
“The historian was only being used for compliance,” said Calva. “It basically worked in the background, and we weren’t getting anything out of it from an operations perspective. But we knew the potential was there. We wanted something for production that could help us discover the ‘art of the possible’ in our operations.”
Multifacility Rollout
Milk Specialties drew up plans to begin rolling out manufacturing intelligence systems across its three primary, human-nutrition facilities. These included the Visalia location, as well as facilities in Fond du Lac, Wisconsin, and Mountain Lake, Minnesota.
They chose the FactoryTalk® VantagePoint® EMI enterprise manufacturing intelligence software from Rockwell Automation, which can integrate a great variety of production and business data into a single information-management, analytics and decision-support system. Within Milk Specialties’ operations, the software collects, correlates and displays real-time process, business and laboratory data.
The system was first implemented at the Fond du Lac facility, In the process of rolling out the EMI software a second time at the Visalia facility, the scope grew as users kept discovering new opportunities for using it.
“An idea for using the software to create in-plant scoreboards came about during the implementation,” Calva said. “Our chief operating officer was the driving force behind that. He wanted everyone on the plant floor to have visibility – to always know if we are winning or losing based on our performance data – and then have the tools to dig into the issue, investigate the root cause of underperformance and fix it.”
At the Visalia facility, tailored scoreboards were placed in three key production areas: separation, filtration, and drying. A fourth scoreboard, displaying plantwide operations, also was placed in the facility’s front office.
The scoreboard in each production area displays plant-summary data, such as throughput, as well as data relevant to workers in that specific location. For example, workers in the separation area can view the fat volume in the skim stream or the temperature in the pasteurizing process, while workers in the dryer area can view evaporation levels and outputs to the bagging line.
Justifying Improvements, Empowering People
“In our industry, production is dynamic and commodity-driven,” said Calva. “We have to be agile, and that means we need more than data. To be flexible in a dynamic environment and respond quickly to change, we need information.”
The EMI software has established a “system of record” that produces a consistent set of facts upon which decisions can be made. This has helped justify operational changes and investments that are crucial to helping the company keep up with market demands.
For instance, process engineers at the Fond du Lac facility wanted to add a new mid-wash process, performing an additional flush between batch runs. The new system showed significant yield improvements following the change. “I was able to use the system to show engineers and management the impact of their change produced a significant increase in throughput,” Calva said. “We are using the same report to measure the efficiency of a costly filter membrane. This has allowed us to stay at peak throughput, without over changing the membrane and adding unnecessary costs. VantagePoint allows us to quantify these policies and initiatives that we decide at an operational level, and then come back after the fact to show how well we’re adhering to them.”
The software also helped resolve an issue with the dryer in the Visalia plant. After reviewing data on the operators’ activities, including seeing down to the level of operators’ sequence of actions. The team noticed that operators were not properly starting up and shutting down the equipment in the right sequence. They relayed this information back to the operators, who corrected the issue.
New analytics from the manufacturing intelligence system also played a vital role in helping the Visalia plant manage its water quality and usage under new California drought regulations. Emergency conservation regulations required state businesses to reduce water usage by 30 percent. Using the manufacturing intelligence system to monitor process metrics, drain flows and water spikes, workers were able to collaborate, using the same information set to keep tabs on water quality and address issues to meet the required reduction, while maintaining quality of waste water.
Meanwhile, the software continues to be put to use in new ways, such as to replace the outdated CIP chemical tracking.
“We’re in the process of building the CIP tracking in the software,” Calva said. “This will save us time compared to the manual tracking and allow us to look at data in new ways to optimize our CIP chemical usage. We’ll be able to compare system-wide against plant-area chemical usage, or standard against manual-override usage. And we’ll be able to drill down to see the exact ounces of chemical used per ton of product produced. This solution is helping drive operations strategy,” Calva added. “It’s our portal to understanding where we are, how we got there, where we should be and what needs to be done to get there.”
The Milk Specialties team is in the process of implementing a manufacturing intelligence system at the Mountain Lake plant. Once the three-facility rollout is complete, they will consider expanding the software to the company’s six other specialty and animal-nutrition facilities.
