Agriculture plays a key role in global food security. Agriculture is critical to global food security and economic development. Precision farming using machine learning (ML) and the Internet of Things (IoT) is a promising approach to increasing crop productivity and optimizing resource use. This paper presents an integrated crop and fertilizer recommendation system aimed at optimizing agricultural practices in Rwanda. The system is built on two predictive models: a machine learning model for crop recommendations and a rule-based fertilization recommendation model. The crop recommendation system is based on a neural network model trained on a dataset of major Rwandan crops and their key growth parameters such as nitrogen, phosphorus, potassium levels, and soil pH. The fertilizer recommendation system uses a rule-based approach to provide personalized fertilizer recommendations based on pre-compiled tables. The proposed prediction model achieves 97% accuracy. The study makes a significant contribution to the field of precision agriculture by providing decision support tools that combine artificial intelligence and domain knowledge.

Christine Musanase, Anthony Vodacek, Damien Hanyurwimfura, Alfred Uwitonze, Aloys Fashaho, and Adrien Turamyemyirijuru. International Journal of Fuzzy Logic and Intelligent Systems 2023;23:214-28. https://doi.org/10.5391/IJFIS.2023.23.2.214

Artificial Intelligence (AI) is increasingly transforming construction processes, yet limited empirical evidence exists on the skills required by professionals across different stages of the construction project lifecycle, particularly in developing economies. This study evaluates the perceived demand for AI skills across the eight stages of the Royal Institute of British Architects (RIBA) Plan of Work in the Rwandan construction industry. A quantitative cross-sectional survey was conducted among 174 construction professionals representing architecture, engineering, quantity surveying and construction management. Data were collected using a structured questionnaire measuring the perceived demand for AI-related skills across the eight lifecycle stages. Descriptive statistics were used to assess and rank skill demand, while the Kruskal-Wallis H test examined differences among professional groups, and the Friedman test assessed differences across lifecycle stages. The findings show moderate-to-high demand for AI skills across all lifecycle stages. The use stage recorded the highest overall demand, followed by strategic definition, and manufacturing and construction, while handover recorded the lowest. Significant differences across lifecycle stages were identified using the Friedman test (χ² = 19.162, p = 0.008), although the practical effect was very weak (Kendall's W = 0.016), indicating modest variation in relative intensity. Most individual skills did not differ significantly across professional groups, suggesting a substantial common set of AI competencies alongside selected discipline-specific requirements. The study’s findings provide evidence for the development of targeted, stage-specific construction education, professional development and workforce strategies in Rwanda and comparable developing economies.