CONA Services Embraces Supply Chain AI

CONA Services LLC is an IT services company owned and governed by the 11 largest Coca-Cola bottlers in North America. CONA is a strategic partner that provides its bottlers with a common set of processes, data standards, and technology platforms. CONA, an abbreviation for Coke One North America (CONA), solutions have been rolled out to more than 500 locations in the US and Canada, representing 80,000 employees and more than $21 billion in annual revenue.

CONA hosts the bottlers' solutions in a Microsoft Azure public cloud environment. The main applications include SAP for backend transactions, Blue Yonder for supply chain management, and Salesforce for sales. Snowflake is their BI environment.

The history of both CONA and the bottlers they support is interesting. In 2009, the Coca-Cola Company unveiled its “2020 Vision and Roadmap for Winning Together.” This vision encompassed both Coca-Cola and their bottlers. The CEO at the time, said that by "working hand-in-hand” with their bottling partners they would build a “unified and aligned system equipped for long-term sustainable growth.”

According to Baron Jordan, the chief product officer for supply chain at CONA, the map of bottlers' territories looked like gerrymandered election districts. In 2020, Coca-Cola acquired Coca-Cola Enterprises, the largest bottler in North America. Other acquisitions and new bottler agreements followed. The result was a streamlined regional bottling structure—one bottler supported the Southeast, another the Southwest, and so forth.

In terms of production and supply chain coverage, these 11 entities work together. While they are separate legal entities, part of the agreement with the Coca-Cola Company was that if a bottler bought one of the new territories, they were required to come on to a common data platform with common data standards. CONA emerged in 2016 to support this vision.

This IT environment is cost-efficient and has very high service levels. It is also best in class in terms of preventing unplanned downtime.

But more importantly, by having all the bottlers on a common platform, the bottlers can work together to meet customer demand efficiently. In short, the 50 production centers owned by the North American bottlers work cooperatively as a common production system to drive supply chain efficiencies.

A group called National Product Supply, with members from all the bottlers, develops the collaborative supply chain plan based on a defined governance processes. The supply chain plans are created across different time horizons—annual, quarterly, monthly, weekly, and even daily. Demand planning encompasses trade sales, projected sales to retailers, and a forecast of nontrade sales, the stock keeping units that will be produced by one bottler on behalf of another.

The demand, supply, transportation, and warehousing plans are created on the Blue Yonder platform. Eventually, these longer-term plans are executed. The production plan is fed into the MRP for production execution. Daily transportation and warehouse plans are developed that go down to the level of what will be picked, packed, and shipped.

Mr. Jordan describes the Blue Yonder user interface as being very user-friendly. “We have dashboards on top of dashboards.” There are green, yellow, and red icons that help users see if things are proceeding on schedule. “We have many types of rules and tolerances and variances that are electronically detected as this data flows through the system.” The system can detect a deviation from a forecast, for example, and yet understand that the deviation is in an allowable range and that an alert does not have to be generated.

However, unexpected events do happen that the system can’t fully handle. For example, a large customer may place a large, unforeseen order that becomes visible at 9:00 a.m. Business operations may decide that the parameters must be changed to support the order. “And we're going to be there with them partnering on how to do that, how to change the data, change the process, and move things forward.” So, while, as an IT services firm, CONA is not responsible for creating the plans, Mr. Jordan points out that “we're deeply involved in daily operations.”

One of the things that Mr. Jordan is excited about is artificial intelligence. Because they have the foundation in place - all their core applications are up to date on the newest release and share common data, the bottlers are in a place where they can leverage “all of this data that we've amassed.” We can take things to the “next level in terms of interoperability and connected planning.” AI will allow the extended enterprise to further break down the silos between logistics and production execution, create more visibility, and thus be “more nimble and more proactive in terms of how we run the entire enterprise.”

One example of breaking down silos would involve using sensor data from production equipment to predict equipment failures before they occur. This would allow for better production scheduling and the ability to produce what was planned with a higher degree of accuracy.

One of the bottlers is also piloting Blue Yonder’s cognitive demand planning capabilities. This solution goes beyond a statistical baseline forecast to an improved forecast model capable based on including other forms of data and using other algorithms. They are looking for the right use case. Should it be used to forecast a group of materials? Specific products? An important customer’s demand? Or for promotions? They have an internal data science team working with their beta bottler on this.

CONA is a big believer in moving away from a “black-box solution.” A black box solution refers to a system or device whose internal operations or algorithms are not readily visible or understandable to those using or interacting with it. “That's not going to work for our business,” Mr. Jordan said. Operations need to understand and know what's going on, and they also want to merge their models with Blue Yonders’ baseline model. Mr. Jordan is glad the Blue Yonder now supports a mix-and-match AI strategy.

Mr. Jordan concluded by saying, “the speed with which Blue Yonder is developing this cognitive platform is remarkable. That doesn't mean that we're going to get onto it without doing our due diligence. And I also must construct the value proposition and business cases. But we're engaged in that. There's a lot to evaluate.”