AI and ML as opportunities to improve insurance penetration in Africa: Case of small scale farmers in Kenya and Ghana

Unfavourable weather conditions and natural disasters pose a significant risk to the predictability of income for farmers. Weather index-based crop insurance based on big data from farmers and satellite, a practical solution is possible in enhancing the resilience of farmers to weather related shocks thereby enhancing food security. Weather index-based insurance is less expensive to administer compared to traditional insurance hence more affordable contracts and faster payments to farmers, who often need the funds for timely planting in the subsequent season. The research aims to enhance de-risking small scale farmers by reducing the costs associated with crop insurance, thereby increasing the uptake of the product. This is achieved using AI and ML algorithms applied on the available big data.

In situ weather stations, remote sensing satellites and citizen data collected via mobile application that captures farmers voice on relevant data including, rainfall, temperature, soil moisture, yield and pests. These data will be used to create a weather index. Machine learning (ML) models will predict the data variables of a specific region based on historical rainfall and soil moisture data going back forty years. The generated prediction is used to determine if the growing period for a farmer will result in a trigger. The trigger is based on historical data where the weather variables are below an established trigger. Claim payments are based on the realization of an objectively measured weather variable that is correlated with production losses.

Crop insurance will be a basic instrument for maintaining stability in farm income, through promoting AI and ML, encouraging investment, and increasing credit flow in small scale farming. Therefore, self-reliant farmers will take up insurance and in cases of crop loss there is claim compensation as matter of right. Digifarm (Kenya) and Farmerline (Ghana) farmers will integrate the ML models in existing platform in addition to voice to provide data and get advice on insurance premiums and claims.

Find the Q&A here: Q&A on 'Artificial Intelligence and Machine Learning'