In the extrusion process for making corn flips, corn kernels are first ground into fine corn grits. This grit is fed through a hopper into the extruder where it is mixed with water to form a dough, see illustration.
The extruder has a long screw (similar to a thread) that rotates inside a cylinder, heating the dough and forcing it through the cylinder toward the outlet. There, under heat and pressure, the dough mixture is forced through a forming die, cooled, cut, and processed. In the end, you have a small, light and delicious flip in your hand.
Experienced machine operators are able to detect changes or malfunctions in systems and processes at an early stage. In most cases, intuitive decisions are made that lead to process control adjustments or even process interruptions.
A high degree of experience is required to correctly interpret acoustic stimuli and derive intuitive decisions from them. This cognitive perception and experience cannot be transferred to other employees through education or training, but must be acquired independently.
The aim of the project is to map the auditory perception of experienced machine operators and thereby realize automated, bio-inspired process monitoring during extrusion. The corresponding feasibility study will be carried out using a starch extrusion process as an example on a pilot plant scale.
For this purpose, sensors are attached to the extruder, which acquire acoustic signals during the running extrusion process. Machine Learning (ML) and Artificial Intelligence (AI) will be applied under different aspects.
July 2022 – June 2023