Advancing Agronomy
HyperLayer Data Collection on the Smart Farm Digital farming technologies — machine learning, artificial intelligence and robotics — will play a large role in the future of sustainable agricultural production. However, innovators require comprehensive training datasets (essentially huge collections of data) to build algorithms for data-driven predictions or decisions, but they don’t currently exist. Since Olds College focuses on innovation in the agriculture industry, it eagerly took on the challenge!
Felippe Hoffmann Silva Karp, PhD candidate from McGill University, collecting ground data from Field 15/16.
Introducing the HyperLayer Data Concept — a process where the Olds College Smart Farm compiles, analyzes, and uses virtually every type of agricultural data to support the development of new technologies for Western Canadian producers. The HyperLayer Data Concept collects data layers with geographical coordinates across the 2,800-acre Smart Farm and stores them in a standardized georeferenced format which allows for quick extraction, easy analysis and data sharing. For each field on the Smart Farm, there will be multiple layers of data collected — topographical data, detailed soil nutrient and moisture mapping, multispectral and hyperspectral imagery, yield data — for analyzing, developing and validating new technologies, and building next-generation machine learning algorithms. Data collection for the HyperLayer Data Concept started in 2020 and there’s still years of data to collect; it’s the most labour and time-consuming part of this project. “The Smart Farm has the expertise, technology and capacity to overlay and analyze many layers of geospatial agricultural data that exists in the industry,” explains Dr. Alex Melnitchouck, Chief Technology Officer - Digital Ag at Olds College who is overseeing the technical aspects of the project. “The HyperLayer approach starts with the creation of georeferenced field boundaries, and then layers of data are collected and analyzed within those boundaries. The information is stored in a geospatial database and used for integrated analysis and predictive analytics on an on-going basis.”
8 Olds College Horizons