Published in Scientific Papers. Series "Management, Economic Engineering in Agriculture and rural development", Vol. 24 ISSUE 2
Written by Mădălina Andreea BACIU, Ștefan VIZITEU, Stejărel BREZULEANU

The plot quality rating of seed research industry is of great importance for most trials and experiments, trying to bring data to validate or invalidate a trait, or to see how a trait is getting variability. Still today, the plot quality is a manual assessment, requires a lot of time, and the decision based need to be taken in a short time for creating space for actions. There is a switch of plot quality rating assessment, from conventional to modern, based on imaging. Many multinational companies who business in agriculture, mostly in breeding and research and development, are now, more and more attracted by the drone use as a proof of AI (artificial intelligence). The importance of a plot quality rating is high, from the pure reason it participates in the data validation, standardization, and governance as a science decision. In this area of research, imaging can be very useful and having strong benefits for the final customers. This need provided by AI comes with the imaging concept, based on drone flights and interpretation. The drone is most of the time a commercial one, but the sensor itself is multispectral, able to deliver different layers with interest for the farm. The farm decision or the research team, evaluate the plots by using a application developed by Sentera (USA), named Field Agent. This app represents today the most advanced plot evaluation, and the use of it is quite simple, most of the time deductible for the professional agronomist. As a POC (Proof of Concept), we have used the AI to bring more reliability into the data at farm level, specifically for corn seed production, where we were able to have a stand count (for female and male), tassel identification (target is to develop it for female rows, detasseled), and yield prediction together with disease and pathogens control under the Power Distribution Board (PBD).

[Read full article] [Citation]