How Data may Improve Farming
The Case / Scenario
The agriculture industry needs to grow its production levels by 70% to feed the world’s growing population, which will exceed 9 billion by 2050. Yet the industry is also faced with reducing its significant ecological impact, which contributes to climate change. Agricultural Smart Information Systems (SIS), the combination of big data and artificial intelligence, is a promising tool to meet these challenges. The Crop Protection Division at BASF uses SIS through their Maglis project to provide farmers with local weather predictions, farm efficiency and sustainability metrics, as well as early detection systems for weed, pests and disease. This personalized information allows farmers to use their land more effectively, resulting in cost reductions, quick and effective crop forecasting, and improved decision-making and efficiency. Maglis consists of three components: (1) information on local weather predictions, plant disease detection, and recommendation of tools to minimise risk; (2) crop and yield previews, farm efficiency and sustainability metrics, and early detection systems for weed, pests and disease; and (3) support system for farming consultants who are working with farmers to find ideal agricultural planning solutions.
Ethical issues identified in the case study largely matched those found in the literature, particularly:
- Accuracy of Data and Recommendations: not all farmers have available data and data retrieved from third parties may not be accurate.
- Data Ownership and Intellectual Property: Farmers worry that giving up their data will make them too dependent on third parties offering data analytics services.
- Privacy and Security: Farmers give up personal information which they fear may be used again them. Drones may intrude on their privacy by collecting data from the air.
One additional ethical concern was identified in the case study:
- Employment: As SIS performs the work of agronomists, there are fears that SIS will replace human jobs in agriculture.
There are several measures taken to tackle the ethical concerns, notably:
- Accuracy of Data and Recommendations: Farmers are given the opportunity to update projections of growth predictions in case of inaccurate data.
- Data Ownership and Intellectual Property: No individual data is used to make comparisons between farms. If this would be the case in the future, explicit informed consent of the farmers is required.
- Privacy and Security: Emphasis is on securing all servers and fully encrypting software to minimize risk that farmer’s data is breached.
- Employment: The SIS technology is merely a complement to rather than a replacement of human workers: the technology is, currently at least, limited in its abilities.
The analysis of the Maglis Project gives insight in how an agribusiness is responding to the ethical implications of SIS technologies. The study supports concerns found in the literature and highlights an additional concern. Important questions relating to the integration of agricultural SIS are: how will it develop and what types of data will be required, and will the farmer have to pay for this service in the future? Specific insights arising from this case study include:
- Accuracy of Data and Recommendations: Who is held responsible for inaccurate recommendations? The company or the farmer?
- Data Ownership and Intellectual Property: There needs to be a mutually beneficial relationship between the farmer and the agribusiness in order to obtain and use data to improve SIS.
- Privacy and Security: Security of platforms is a priority to safeguard farmers’ privacy and livelihoods.
- Employment: SIS technology is fallible, and farmers may value advice of a human agronomist over AI-created prescriptions, as farmers do not fully trust recommendations given by AI tools.