Ultrasonic Sensors and Machine Learning for Digital Food and Drink Manufacturing

The world is experiencing the 4th industrial revolution which involves the use of digital technologies such as artificial intelligence, cloud computing, sensors and the industrial internet of things. These digital manufacturing technologies have the potential to improve manufacturing productivity and efficiency whilst reducing the environmental impact it has. Food and drink is the largest manufacturing sector in the UK, contributing almost £30bn to the economy every year. One barrier preventing the widespread adoption of digital technologies within food and drink manufacturing is the lack of suitable online technologies capable of measuring the properties and therefore quality of the food.

Ultrasonic techniques use mechanical waves to probe and therefore characterise the properties of multicomponent materials (e.g. food) and are an attractive sensing technology due to their low cost and size. However, for them to be a suitable online sensor, new signal and data processing algorithms are required to relate sensor measurements to the food’s physical properties. Machine learning is a form of AI, which uses vast amounts of data to develop predictive algorithms. A key advantage of machine learning techniques is the variety of data they can process and their ability to improve as more or better data becomes available.

This project will focus on developing ultrasonic techniques, which utilise machine-learning algorithms to classify the structure and quality of food. The student will work with industrial partners focussing on solid food products within the bakery, confectionary and prepared fruit and vegetable sectors.