How can robotics accelerate African agricultural autonomy

Abstract

Even in the increasingly globalised world, for African states to fulfil their true potential, they must aim to be fully autonomous and self-sufficient in many areas, chief of these being in agricultural production. Africa not only has the highest percentage of arable land, more than half of its population is employed directly in the agricultural sector, but also has agriculture as its biggest contributor to its Gross Domestic Product (GDP). The future productivity of the African agricultural sector and its ability to maintain food security for its growing population is quite closely tied to its ability to fully harness contemporary tools and technology available. By providing valuable insights into farmlands, these tools and techniques empower farmers to make informed decisions regarding the health of their farms/anticipated yields while also identifying areas of concern and long-term trends.

Current access to these insights is however limited to those who can afford both the sophisticated equipment, such as drones, and the expertise required to operate them safely and process the data outputs. In this work, we present an alternative view based on three years of fundamental research in the African Robotics Unit (ARU) at the University of Cape Town that seeks to reduce the barrier to obtaining these insights. We do so by reinterpreting the task of obtaining both geometric and semantic insights in a FARM as a fundamental robotic SLAM problem allowing current state-of-the-art techniques developed within the African Robotics Unit to be leveraged. In this way, by empowering more farmers with more information about their local context, they are able to make better decisions and thus improve the productivity of the collective sector, accelerating African agricultural autonomy.