Artificial Intelligence- Bridging the data gap in farming
By Derek Greene
In the past, Artificial Intelligence (AI) was largely confined to the fields of Computer Science and Statistics. Increasingly these methods are being adopted in a range of industries and sectors, from financial services to medical diagnostics. However, for some applications, the lack of relevant data for building and evaluating AI models creates a barrier to adoption.
In contrast, the availability of rich, large-scale datasets in the agricultural sector makes it a potentially rich area for employing AI. Individual farms produce substantial volumes of data on a daily basis, which will increase as low-cost sensors are adopted, leading to the constant collection of continuous “streams” of data. This data is heterogeneous, ranging from measurements of soil conditions to infrared spectral analyses of milk. Given this rich data, the challenges for researchers are twofold. Firstly, how can we integrate these diverse datasets at a national level in a practical way? Secondly, what are the appropriate AI techniques to use for modeling the integrated data, in order to help farmers and processors make the right decisions? A variety of projects in VistaMilk are already being undertaken to answers those questions. These range from developing new algorithms to predict grass growth to support sustainable dairy farming, to building deep learning architectures that automatically classify agricultural land use based on satellite or drone imagery. Working with VistaMilk industry partners, researchers aim to not only develop novel algorithms but also provide decision support technologies that will have a real impact in Irish farms.
For more information contact; Derek Green email@example.com