Milk Specialties Global: EMI Scoreboards and a 30% Water Target Met

Milk Specialties Global, a Minnesota-based maker of milk, whey and specialty proteins with 10 plants, was growing quickly. Rockwell Automation's October 4, 2016 case study says its production control systems were data silos with little ability to track or compare data over time, and its reporting was outdated and labor intensive. Tracking clean-in-place (CIP) chemical use meant downloading machine data into a Word document and re-entering it in a spreadsheet to build a graph. A historian inherited with the Visalia, California plant was used only for compliance. Rockwell has since removed the page; this summary follows the Internet Archive copy captured on January 20, 2026. The source gives no baseline throughput or yield.

Milk Specialties chose FactoryTalk VantagePoint, Rockwell Automation's enterprise manufacturing intelligence (EMI) software, which Rockwell Automation says collects, correlates and displays real-time process, business and laboratory data. It went first into the Fond du Lac, Wisconsin plant, then Visalia, with Mountain Lake, Minnesota under way. At Visalia, scoreboards in separation, filtration and drying show plant-summary data such as throughput plus local data, for example fat volume in the skim stream or evaporation levels in the dryer. A fourth scoreboard shows plantwide operations in the front office. The same system was used to review a mid-wash step, a filter-membrane policy, dryer operator sequences and water use. The account describes data collection, reporting and dashboards. It does not describe machine learning or predictive models.

Rockwell Automation reports these outcomes:

• Throughput: at Fond du Lac, the system showed a significant throughput increase after an extra flush was added between batch runs. No percentage is given.
• Water: at Visalia, monitoring process metrics, drain flows and water spikes helped workers meet California's emergency 30 percent water-use reduction while keeping waste-water quality.
• Dryer start-up: operator-action data showed equipment was started and shut down in the wrong sequence, and operators corrected it.
• Reporting: Rockwell lists automated reporting as saving time over manual reporting. CIP tracking was still being built when the account was written.

This is a vendor-published 2016 account quoting plant staff. It gives no throughput, cost, water or time-saving figures beyond the 30 percent requirement.

Bring these checks to an EMI demonstration, and run them on your own data:

• Ask for the before-and-after data behind "significant increase in throughput": which report, which period, and what else changed. The source gives no figure.
• Ask which of your systems the software must connect to. Milk Specialties started from control-system data silos and a compliance-only historian, so check how process, business and lab data are correlated, who maintains the tags, and how bad data shows up on a scoreboard.
• Test the action loop. The dryer and water examples worked because people acted on what the data showed. Ask who owns a missed scoreboard target, how an investigation starts, and what licensing and rollout costs each added plant brings; the source gives none.

For related guidance, see the manufacturing data analytics guide and other Food & Beverage deployments.