Designing AI for food justice and sustainability

Publié le mercredi 31 mars 2021

Cet article de réflexion est uniquement disponible en anglais pour le moment.

Emergent agricultural innovations in big data and AI have the potential to revolutionize food production for the better – making it easier, more productive, and more sustainable. Yet, deploying these technologies in an equitable way remains a challenge. The Alex Trebek Forum for Dialogue’s Project on AI for Healthy Humans and Environments at the AI + Society Initiative conveyed a conversation on how to ensure that the so-called “digital revolution in agriculture” is both equitable and sustainable. Designed as a dialogue between academia, industry, and policymakers, the panel featured Dr. Kelly Bronson (Canada Research Chair in Science and Society, University of Ottawa), Dr. Ranveer Chandra (Chief Scientist, Microsoft Azure) and Giuliano Tolusso (Deputy Director, Innovation and Growth Policy Division, Agriculture and Agri-Food Canada).

Digital technologies are being adopted across the agri-food system. This panel was designed to highlight the impacts of these changes from the points of view of the industry, academia, and government. Dr. Kelly Bronson showcased work that intends to amplify voices of smaller-scale and“alternative” farmers (e.g. organic growers), who are being overlooked in the transition towards so-called “smart” farming. Dr. Ranveer Chandra explains how big data and sophisticated computing can augment smallholder farmers’ knowledge of the land with insights that increase profitability and sustainability. Finally, Giuliano Tolusso presented a perspective from the government of Canada which is working towards increasing the adoption of digital agricultural technologies in an inclusive and sustainable way.


Key insights

Dr. Kelly Bronson presented work she is doing with an interdisciplinary research team including engineers, data scientists, farmers, and social scientists. The project is funded by competitive SSHRC Insight Grant and is called “Diversity by Design: Engaging small, unconventional growers in the smart farming revolution.” Dr. Bronson began the panel by explaining that companies like John Deere and Bayer-Monsanto collect agricultural data in much the same way that Facebook and Google collect personal data. Aggregated agricultural data is then fed into algorithms and analyzed for the purposes of acquiring insights about farm operations, which then, in turn, are provided to the farmer. These insights can provide farmers with new information about their operation that might alter their behaviour and practices on their farm, ideally facilitating environmentally best management practices Dr. Bronson’s project rests on the issue that most of the agricultural data collection and analytics are biased towards large-scale, conventional, commodity crop farms. Farmers that operate mixed crop operations are less likely to adopt these technologies because they are not built to serve those farming processes. Dr. Bronson’s work aims to fill this gap in an “innovation ecosystem” that leaves out “sustainable producers'' such as farmers engaging in regenerative, organic, or agroecological practices. Dr. Bronson explained how her project will involve participatory workshops that engage smaller-scale farmers in the prototyping of digital agricultural technologies. Dr. Bronson contends that it is imperative to amplify the voices of smaller-scale farmers, who are crucial in terms of increasing the ecological resilience of our food system.

Dr. Ranveer Chandra, discussed complementary work that is going on at Microsoft. He’s a computer scientist that has been very involved with the establishment of a program at Microsoft called FarmBeats. One of the goals of this project is to enable data-driven agriculture, particularly on small-scale farms like the one his grandparents used to farm in India. He believes the innovations being deployed through FarmBeats can help smallholder farmers have access to the same technologies that are currently being used on large, more capital-intensive farms. Dr. Chandra found that there are two main challenges that are barriers to adoption for many farmers across the world: low internet connectivity and lack of affordability. One of the FarmBeats innovations addresses the challenge of low internet connectivity in farm fields by using unused TV channels to send and receive data. Another FarmBeats technology allows farmers to use their smartphones as aerial sensors. By mounting a smartphone on a weather-proof mount on a helium-filled balloon, farmers can then have access to remote sensing data that may not have been available to them before. While soil sensors and drones might be expensive and inaccessible to smallholder farmers in the developing world, most of them still have smartphones and would easily be able to deploy this type of technology. Dr. Chandra and the FarmBeats team believe that driving down the cost is one way to democratize access to digital technologies that are otherwise far too expensive for farmers in the developing world.

Finally, Giuliano Tolusso explained how the Agriculture and Agri-Food Canada (AAFC)’s interest in precision agriculture aligns with the work of both Dr. Bronson and Dr. Chandra. His department looks at how policy can facilitate the adoption of precision agriculture technologies while taking issues of sustainability, productivity, competitiveness, and equity into consideration. Tolusso indicated that AAFC has taken account of Dr. Bronson’s research showing that digital technologies are more likely to be adopted by capital-intensive farms in Western Canada, rather than smaller-scale farms in Eastern Canada; in particular, AAFC has been working to support the democratization of access to precision agriculture technologies through participation in a federal government program called the Innovative Solutions Canada program. AAFC has posted a “challenge” that is focused on the “scaling down” of precision agriculture technologies so that they can be used more effectively by smaller-scale farms (e.g. cost-effective and open source tools). Among the successful applications was an ‘Internet of Things Smart Data Hub” which is compatible with any farm management tool, and ‘AI-enabled autonomous drones for fully automated crop protection’ used for pesticide application. AAFC is also investing money to support agricultural discovery science and innovation through its Living Laboratories Initiative (LLI), which was launched in 2018. This project is an integrated approach to innovation that brings together farmers, scientists, and other stakeholders to test and monitor best management practices and new technologies in farming environments.  It was mentioned that Dr. Bronson was also collaborating with AAFC on this living labs initiative as an Advisor to the LLI and through a separate SSHRC-funded project of hers.

During the conversation period moderated by Alex Trebek Postdoctoral Fellow Dr. Mascha Gugganig, panelists discuss the ways in which AI and Big Data technologies are being deployed by urban farmers, especially in vertical farming environments. The panelists also had a chance to discuss one of the most prevalent barriers to the adoption of these technologies which is a lack of trust in data governance frameworks. While there are several initiatives working to make data governance fairer, like the voluntary certification Ag Data Transparent, trust remains low. Furthermore, all panelists asserted their belief that ethical AI principles should be applied to data analytics in agriculture. Biased datasets could be harmful, not only to farmers but to stakeholders all along the supply chain, including consumers. In conclusion, all panelists agreed that despite the fact that digital agriculture is promising in many ways, researchers, technologists, and policy-makers alike must include diverse voices and attend to the social and ethical dimensions of these emergent tools.


Watch the event


Key resources to learn more


Our event summaries are provided to help amplify the conversation around the ethical, legal, and societal implications of AI in a short and accessible way. We invite you to watch the video and read the additional resources for more information on this topic.

This summary was prepared by Sarah Marquis, a PhD student at the Institute of the Environment, with the Canada Research Chair in Science and Society. Opinions and errors are those of the authors, and not of the Initiative or the University of Ottawa.

Haut de page