The technologies driving this include sensors, robotics, the industrial Internet of Things, additive manufacturing, artificial intelligence, virtual and augmented reality, digital twins and cloud computing. Although these technologies can be used standalone, the true benefits are reaped when used in combination with one another.
For instance, Nik Watson, associate professor at the University of Nottingham, has been undertaking research within food and drink manufacturing focused on using non-invasive sensor technology in combination with machine learning and cloud computing.
Watson said: “With the industry partners we’ve worked with, sensors have been integrated into devices that are then connected to an industrial Internet of Things cloud platform that streams data directly to our research team’s laptops in the lab. By adding machine learning and smart algorithms we can then analyse that data and develop outputs. We call this generated actionable information which enables firms to make informed decisions about the production environment.”
One of these projects saw the University of Nottingham working with Loughborough University to develop an intelligent multi-sensor technology to monitor the removal of surface fouling during cleaning of processing equipment.
Watson said: “The job of autonomous clean-in-place processes is to clean away any fouling materials which would contaminate the next production batch and affect process efficiency, but they often over-clean, which has a significant impact not only economically but environmentally too. Using simple ultrasonic and optical sensors we have monitored these processes in real time to determine when the fouling had been removed during cleaning. Using these measurements from the sensors, several machine-learning models were also studied to predict the presence of fouling.”
Another project came through the University of Nottingham’s involvement in Connected Everything, an Engineering and Physical Sciences Research Council-funded research network. As part of its remit of supporting manufacturing it has funded feasibility studies. One of these was Brewnet, a study into how cloud-connected sensors would enable small-scale process optimisation at brewer Totally Brewed.
“We used ultrasonic and temperature measurements and cloud-based predictive analytics to predict the alcohol-by-volume percentage (ABV %) during fermentation,” said Watson. “For this we developed an ultrasonic sensor, in a bespoke probe holder, to measure alcohol content in a 200-litre fermenter. With connectivity between the sensing platform and a cloud platform the data could be stored and processed. Signals were sent between the brewery and the University of Nottingham via low-power wireless networks.”
Less beer wasted
Key findings of the study included, firstly, that ultrasonic measurements can be combined with machine learning to accurately predict alcohol content during fermentation and, secondly, that off-the-shelf technologies can be used to develop a cloud-connected sensor system.
Additionally, reducing over-fermentation and eliminating the need to remove beer to measure the ABV % offline led to an economic benefit for the brewer and it reduced the amount of wasted beer and resources.
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.