
Expansion of the capabilities of the AI module of our system
To date, we were using the AI module included in our most advanced systems solely for estimating the external quality of the fruit.
Given the good performance of this work strategy based on AI, we decided to assess the application of this tool also for fruit measurement, especially in those cases in which it is difficult to obtain the equatorial diameter of the fruit when there are no other related measurements that allow detect its position without errors.
We are going to carry out a first application of the tool with a specific training that will seek to label each fruit with the value of the equatorial diameter.
Although the task of obtaining the working dataset is going to be quite expensive at first, this new approach could allow the simplification of the mechanical part of the machines since the strategies we use for machine learning are highly tolerant to changing situations. such as a perfect alignment of the fruit or a controlled rotation.
In the same way, during 2023 we are going to start expanding the external quality module for other fruits for which we currently do not have applications, such as avocados and mangoes.