2020 - OSIsoft PI World San Francisco - Life Sciences, Food, Beverage
Deschutes Brewery Uses Predictive Analytics to Delay Glycol Capacity Expansion
Deschutes Brewery has been fermenting iconic craft beers in Bend, OR since 1988. Over the past several years, Deschutes has worked to create predictive models that use past fermentation data to predict key points in current fermentations. These models were initially developed to reduce manual measurements and increase product consistency, but late in 2018, another opportunity for these models became apparent. Deschutes was going to have to purchase a new chiller to have enough cooling capacity to continue running their current process, but if they could use the fermentation predictions to naturally stagger fermentation cooling steps, they could drastically decrease their maximum chiller load and delay the purchase of a $750,000 chiller. To do this not only would the predictions have to be working well, but operations would have to be convinced of this and begin using them regularly.
Kyle Kotaich is the Senior Data Analyst for Deschutes Brewery; one of the largest family and employee owned craft breweries in the Pacific Northwest. With the ambition to make better beer more efficiently, he has been the leading developer of the PI System, analytics, and machine learning for brewing and packaging at Deschutes Brewery.
Tim Alexander is the Operations Technology Manager at Deschutes Brewery, one of the largest family and employee owned craft breweries in the Pacific Northwest. Over the last 13 years at Deschutes Tim has been the catalyst of digital transformation at the brewery. From development of SCADA systems to integration of machine learning, his insight and experience has guided nearly all facets of technology at Deschutes Brewery.