Effective Plantation Management with Crowd-sensing and Data-driven Insights: A Case Study on Tea

We have an assortment of upcoming technologydriven ways to assist specific activities with precision such as spraying with drones, pest assessment with images, mechanisation to complement farm labour, macro farm health assessment with remote sensing, and Internet of Things (IoT) in general that plays a key role on the farm-edge in different capacities. With effective digital plantation management as an end-objective, we present our work on development of the framework constructs to (a) digitise pest management activities to record crop-stress data along with field operations, and (b) build insights from the data to respond faster to stress incidents with precise control measures. As part of the digitisation, we employed design thinking concepts and a human-centric approach to develop user-friendly interfaces where crowd-sensing with the help of ground staff is used as a foundational activity. Descriptive and diagnostic insights on the gathered data were brought out to correlate incidents with operations based on aggregated patterns, and generate deep insights on crop images with artificial intelligence. Image-based insights include localisation and recognition of symptoms associated with insect pests, diseases, and nutrient deficiencies that were non-trivial to get earlier through manual operations. Such insights were used to generate system recommendations that support experts in issuing effective advisory towards curative action on the field thus sowing the seeds for an Industry 4.0 future for plantations.