Lead Institution: University of Nottingham
Partner Institution: University of Leeds
Industrial Partner: Totally Brewed
Project Team: Dr Nicholas Watson (PI, Nottingham) and Dr Ali Zaidi (Leeds)
Project Duration: 12 months (01 April 2017 – 31 March 2018)

Information Sheet

Research Challenge
The food and drink industry is characterised by high volume production of low value products. SMEs operating within this sector often fall behind with technological advances due to a lack of capital to invest in new assets and limited or non-existent research capabilities. However digital manufacturing does not require expensive new hardware or onsite expertise as process benefits such as reduced costs, resource utilisation and waste are delivered via the collection, analysis and decision support capabilities of data. This data can be collected by low cost sensors which are connected to a cloud server for near real-time predictive analytics. One example of small scale food and drink processing is craft brewing.

A critical stage of the brewing process is fermentation, where yeast is added to the wort to convert sugar to alcohol. The fermentation process is complete once the beer has reached the desired alcohol content. This is currently determined by removing a sample and measuring the specific gravity using a hydrometer. Although the fermentation process duration should be identical for each batch of a particular beer, this is rarely the case due to seasonal variability in ingredient (malts, hops, water) properties and natural fluctuations in process temperature. As specific gravity measurements are only taken every 4-10 hours (or longer if no-one is working overnight) this often leads in over fermentation. This results in an inferior product and an inefficient use of resources such as electricity for heating. Sub-optimal fermentation also has a significant effect on downstream processes (e.g. Kegging) and often results in the need for additional processing steps (e.g. extra sugar addition).

Ultrasonic techniques use high frequency (>100 KHz) low power (<100 mW cm-2) pressure waves to non-invasively measure solid/liquid materials physical chemical properties. Ultrasonic techniques offer a low cost solution to process measurements but currently no commercial solutions exist for an SME brewing environment due to integration challenges with legacy technologies and expertise requirements.

The aim of this multidisciplinary feasibility study is to combine process sensors, wireless networks and cloud based cognitive algorithms to optimise small scale fermentation processes. It aims to demonstrate how emerging digital technologies enable this optimisation at low cost and without the need for onsite instrument and data specialists.

This project will develop an ultrasonic sensor which can measure alcohol content in a small scale beer fermenter. This ultrasonic sensor will be combined with temperature and pH sensors and connected to a cloud empowered gateway. Intelligent algorithms utilising real-time and historical fermentation data will be developed within this environment to predict optimal fermentation end point. All real-time fermentation data will be used to study the performance of the fermenter and ingredients (e.g, yeast) and made openly available to other craft brewers.