Next Generation Delegation 2014 Commentary Series
By Adam Riesselman, PhD Candidate in Bioinformatics and Integrative Genomics at Harvard University and 2014 Next Generation Delegate
As a bioinformaticist, I like to think that biology acts in entirely predictable ways: our only limitations rest on access to information, efficient retrieval, and clever biological insight. Big Agriculture is also betting on technology and data management in the productive fields of the United States. If agricultural inputs can be boiled down to numbers—such as plant yield across fields, soil quality, chemical input, planting rate, and seed variety—more cost-efficient agronomic solutions can be found.
Adding data management practices to agriculture is well-founded: with cereal yield growths plateauing, growers need a new tool to apply to their fields. Platforms such as John Deere's GreenStar, Ag Leader, and Monsanto's FieldScripts promise $5-9 savings per acre by variable rate planting, minding replanting at endrows, crop variety selection per location, and selectively applying chemicals and fertilizer.
Applying this technology to agricultural production, however, can be complicated. With the average age of the US farmer currently at 55 years old, many precision agriculture end users are often left in the dust about the finer details of data management. Moreover, as farmers age, planting is increasingly being performed by drones, such as computers that drive and control planters, combines, and tractors. Moreover, the line between farmers' control and corporate control of agricultural data is being blurred, both in regards to data obtained throughout the growing season as well as the actual mechanical control of the machine.
The abstraction of farming decisions from the grower to a data-driven corporation will become even more widespread in the future. For example, using Monsanto’s FieldScripts program, growers give the company data such as field boundaries, past yield data, and fertility tests. Monsanto then selects the planting rate and crop variety across the field. This information is uploaded onto the computer in the planter, and the machine takes off to automatically plant the field.
Effective data management practices are paramount in these situations to mitigate any disruptions to our food supply. For instance, yield data is uploaded to a digital cloud whenever the tractor is run through the field. Many farmers are unaware of who owns the data produced on their fields and who it can be shared with. Could this data be used to manipulate agricultural markets if yield information is known instantaneously? Monsanto has already experienced a data breach in their precision agriculture system which could lead to further unintentional release of confidential farm information.
Moreover, malware and other deleterious software could be used to seriously disrupt our food supply. Weaknesses in network security can be found in nearly every computer system connected to the internet. With planters, combines, and tractors constantly interconnected with satellite or cell phone transmissions, network security should not be overlooked. Foreign entities could use digital attacks such as Distributed Denial of Service (DDoS), Permanent Denial of Service (PDoS), or viruses to render our planters useless in the spring. Our combines and sprayers could sit idle in the field if mechanical components were fried by malicious code sent through the web.
Though agriculture can be abstracted down to data points on a field map, we must use this technology wisely. Vulnerabilities come with a lack of diversity in a network system. If the network is down, our food production system is at risk as well. Data security has now become food security.