Data Farming: Facilitating sustainable intensification through digital innovation

Rapid growth in the global population puts agriculture under ever-increasing strain. Estimations suggest a 70-100% rise in food production is necessary by 2050. There is an evident need to develop and utilize new and inventive farming techniques. This task, however, is exacerbated by environmental externalities, leading to a steep rise in global temperatures and extremes of weather. The consequences of these issues are disproportionately concentrated in the developing world, where inequality and poverty are rampant. It is the intention of this study to critically appraise existing approaches and suggest pathways to agricultural prosperity through digital innovation.

Sustainable intensification (SI) of agriculture has emerged as a key strategy in climate adaptation, international development and food security. Defined as "a process or system where agricultural yields are increased without adverse environmental impact and without the conversion of additional non-agricultural land", SI provides a comprehensive solution. Technologies include the use of fertilizers, improved crop cultivars, soil and water conservation, intercropping, crop rotation and conservation agriculture. Providing smallholder farmers with greater access to SI technology is perceived as essential to increasing productivity, sustainability and resilience in the global South.

The economic study of adoption strives to understand the transferability and utilization of SI technologies. Several recent studies, however, have identified pervasive problems in the methodology and practice of technological change. Adoption as a concept captures a very limited picture of the decision processes and external factors that influence the uptake of SI. Studies draw from small datasets, with few explanatory variables, without a longitudinal perspective and rely upon a binary vision of success and failure. Due to these data biases, many of those who most require assistance are excluded or unrecognized. Studies do not appeal to the values of participants, obscuring a range of societal factors such as background, political beliefs, 'handed down' farming practices or the relationship of farmers to their wider community. There is an apparent disparity in understanding and awareness between researcher and respondent which requires further exploration.

Digital innovation refers to a broad spectrum of technologies which have in recent years proliferated throughout all aspects of society. Two of the most pervasive have been data analytics in shaping governance and policy, and ICT in mediating social lives and knowledge sharing. Both initiatives have shown great potential in advancing sustainable agriculture. In this case, they present an opportunity for researchers to better understand and model the needs and preferences of farmers and, following this, to develop and implement technologies that are more appealing.

Thus, the focus of this PhD is two-fold: to employ machine learning algorithms to distill a set of key adoption predictors and gain a superior understanding of technological change and the personal data required for its estimation; and to explore the potential of mobile phone-based service provision in developing countries as a key driver of SI technology awareness, understanding, uptake and peer support. It is intended that these measures may bridge the gap between users and researchers, enriching their mutual understanding with more accurate and accessible data.