Rubber farmer assistant: An interactive digital platform for assessing climate variability and soil degradation in Hevea brasiliensis

Abstract

Rubber (Hevea brasiliensis) is a major plantation crop in Sri Lanka, yet the smallholder productivity has dropped due to increasing climate variability and soil erosion. The farmers typically do not have access to place-specific, timely agricultural guidance and hence adopt inappropriate adaptation strategies. The present study aimed to conceptualise and pilot test a digital platform, Rubber Farmer Assistant, for providing science-based, real-time advice for enhancing resilience and productivity. Data collection was informed by the Technology Acceptance Model (TAM) and the Diffusion of Innovations Theory. A mixed methods design was used, combining quantitative (structured, closed-ended questionnaire) and qualitative (open-ended responses) approaches. Quantitative data were collected from 32 smallholder farmers, and usercentred design was informed by qualitative responses derived from open-ended questions. The tool was pre-tested for validity (Cronbach’s α =0.84). Quantitative data were analysed through descriptive statistics using Python. Supporting secondary data from the Rubber Research Institute of Sri Lanka and other institutions in the country assisted in building a robust scientific basis.A central component of the platform is the Farm Health Score Calculator, which provides site-specific recommendations derived from weighted indicators like soil colour, drainage, ideal rainfall (1,500–3,000 mm), and temperature (20–35◦C), with corresponding weights of 26.7%, 26.7%, 26.7%, and 20%. The application has a modular design with provisions for soil and climate analysis, agricultural advice, and expert contact support. Participatory design and iterative farmer testing enhanced usability. Results showed that 81.3% of the respondents had frequent climate problems, and 71.9% had soil erosion. Adoption was enthusiastic, with 93.8% interested in using the platform and 74.2% giving the highest rating to expected improvement (mean = 4.6, SD = 0.72). A chi-square test confirmed a positive correlation between climate impact and adoption of the platform (χ2 = 12.34, p < 0.05). This study confirms that digital agriculture technologies have the potential to bridge information gaps, enable sustainable agriculture, and scale across similar agro-ecological zones.