John Charlie Hoxmeier, Gilded Goat Brewing Company, Fort Collins, CO, U.S.A.
Coauthor(s): John Hoxmeier, Gilded Goat Brewing Company, Fort Collins, CO, U.S.A.; Rick Faulkner,
Fort Collins, CO,
Brewhouse Operations I
Thursday, October 12
The brewing process is well understood but the data and the outcomes are becoming increasingly complex. The demand for specialty beers, style derivatives, and new flavors combined with brewer innovation are driving an exploding need for a flexible brewhouse manufacturing execution system (MES) that supports better analytics for both the brewer and the process. This research revealed that there are more than 200 distinct MES attributes to be tracked, from raw material through recipe management to finished beer. MES is only a component of the information spectrum. The information management landscape for a brewery with a taproom is fragmented and diverse. New beer styles and recipes require fundamental and sound data tracking in order to maintain the foundations in beer production: innovation, consistency, and quality. Most small craft breweries use a variety of home grown applications for MES, including manual or spreadsheet-based applications. Yet the analytic needs of the craft, nano-, or microbrewery are no different than their larger counterparts. In addition, the MES component should provide for some level of integration with the related business systems: supply chain, purchasing, inventory management, taproom management, point of sale, and accounting. This presentation will describe a research project designed to capture the essential data in the brewhouse operation. The process includes analyzing the data attributes of the problem domain of the brewer and present the steps required to design and build a solution domain as the product moves through mash to kettle to fermentation to finish. The ultimate goal was to design a data repository that closely matches the problem domain and yet is flexible enough to serve as a strong ad-hoc analytics platform. The brewing process is decomposed into a conceptual data model. This required a team that represented both brewing and database design expertise. The conceptual model is presented using the entity–relationship (ER) modeling and use case techniques. The model was then normalized so that the data structures would support a relational database implementation. This step is critical for maintaining entity relationships, data integrity between and among attributes, and long-term data quality. Finally, the logical data model is translated into a physical model that is implemented as a cloud-based relational database. During the design and implementation process it is important to understand the semantics of the brewing manufacturing operation. The high-level process is straightforward. A recipe drives the supply of raw materials and dictates the batch plan. Producing a batch consists of several stages with different metrics, inputs, and outputs. The outcome is a finished product that should be repeatable. But the process is not that simple, especially when there are so many potential outcomes by modifying just a couple of ingredients or the parameters. The design process will be described and the finished cloud-based database application will be presented at the conference.
John C. (Charlie) Hoxmeier is the brewmaster at the Gilded Goat Brewing Company in Fort Collins, Colorado. He holds a Ph.D. degree in microbiology from Colorado State University. He is a contracted researcher to the Center for Disease Control and Prevention in Fort Collins. He has been published in several scientific journals.