Research Results

Fruit quality prediction

In the project “Future sensors and digital twins to improve perishable food quality during distribution” an application programming interface (API) is developed to predict fruit quality based on actual recorded storage and transport conditions.

shutterstock_1685648584.jpg

Objective

In perishable supply chains reliable quality prediction can help greatly to prevent food losses. Knowing more about the expected quality upon arrival of a specific a fruit batch, greatly supports in taking more informed decisions to, for instance, adapt settings or change the duration of the chain.

Results

With the developed application programming interface a prediction of fruit quality (remaining storage life) for a fruit batch is calculated, using recorded storage/transport conditions of a certain reefer container and using physiological product characteristics as input. The conditions include recorded temperature, relative humidity, oxygen and carbon dioxide levels over a certain period. The API can also provide warnings e.g. with respect to risks on chilling injury and oxygen or carbon dioxide damage. The interface is available for many different fruit types, amongst others avocado and banana.

Relevance to industry

Nowadays many conditions during storage and transport of perishables are recorded, while the product quality itself is more difficult to monitor through the whole chain, especially in a reefer transport. A quality prediction using specific product parameters in combination with actual recorded storage and transport conditions is a step forward towards more reliable predictions. These are valuable to use as support for decisions in the supply chain.