Deep-Learning
Images of foreign material captured by the TOMRA sorter are analyzed by a deep learning model, reducing the detection of false alarms and equipment downtime. Only true positives trigger alarms to the control system, and those responsible in the company receive instant notifications on their mobile phones and can access the stored images. Additionally, these images help to constantly improve the solution by continuously retraining the deep learning model.
The networked TOMRA Digital Operations Centre
FMID is integrated with TOMRA Insight, a cloud-based platform with an easy-to-use interface and real-time insights. The connected TOMRA Digital Operations Center enables remote monitoring and management of the solution, improving its performance. Since the focus is on sorting and removal of foreign material, the real-time aspect of image processing techniques is elementary. Different sensor technologies, in combination with image processing algorithms, enable proper separation of products and foreign material and thus prevent food waste.
![TOMRA's Foreign Material Identification (FMID) solution was honoured with the International FoodTec Award 2024 in silver.](https://media.koelnmesse.io/aft2021/redaktionell/aft/img/magazin/tomra-foreign/tomraforeign_1200x675_m24_full_m36_1025.jpg)
TOMRA's Foreign Material Identification (FMID) solution was honoured with the International FoodTec Award 2024 in silver. ©TOMRA
Minimise false alarms and food waste
Since traditional methods work in isolation and mostly without connectivity, their potential for continuous improvement is limited. Manual interventions often lead to human error and in addition sorters usually aggressively sort out foreign material to ensure an uncontaminated food product, resulting in a high number of false alarms and also an undesirable amount of good products being lost or having to be reprocessed. The new Foreign Material Identification solution combines the strengths of advanced sorters with the added benefits of a connected, cloud-based solution and artificial intelligence. It reduces the risk of food contamination, improves food safety, and ultimately leads to a safer and more reliable food supply chain with minimized food waste that protects consumers from potential health risks. The information collected in the system ensures the necessary transparency regarding the quality of raw products and sorted products and thus supports internal quality assurance. The versatility and adaptability of the sorting solution allow it to be widely used in many food categories.