Machine learning, big data and the future of digital agriculture

Authors: Jaradat, Abdullah

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 11/13/2019
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Analytics algorithms based on machine learning of big data generated by digital agriculture are transforming the traditional "observe-measure-response" cycle and heralding a new era in agricultural research and development. While, digital agriculture is evolving at a rapid pace into a complex system of increasingly integrated biological, physico-chemical, mechanical, and computational technologies. Meanwhile, production efficiency and sustainable utilization of the production base of futuristic agroecosystems that can feed a burgeoning world population need to be optimized through improved design, precision management, and timely prediction of their responses to external drivers and stressors, with the promise of answering the perennial question of sustainability in agriculture. Long-term agroecosystems research was envisioned as a bridge between sustainable provisioning of ecosystem services and productive utilization of the production base of futuristic agroecosystems, and their interaction with the ecological landscape. Advances in machine learning and big data analytics of multiple data sources will be reviewed and presented within the context of on-going research efforts using integrated stationary and mobile sensors, image classification and segmentation algorithms, deep learning algorithms, and high-dimension statistics within a time-series context. In due course, these technologies are expected to provide farmers with real-time information that can enhance their management capabilities and enable them to make timely on-farm decisions. However, for farmers to embrace the digital transformational trends in agriculture, the hype of digital agriculture may not be appealing to them unless there is a clear value proposition that they can easily and economically integrate into their businesses and earn a reliable return on investment. Nevertheless, from measuring and managing inputs to documenting and quantifying sustainability metrics, farmers who adopt digital tools will have better opportunity to effectively participate in a global food economy that can feed the world, while demanding documented transparency throughout the entire food production cycle.