selemansolar-ai/Tanzanite-Solar-Cell-Designer-AI

AI tool to predict Solar efficiency using Tanzanian minerals.Built for Microsoft Research and IBM Research Africa. # Tanzanite-Solar-Cell-Designer-AI The world's first AI tool to model solar cell efficiency using Tanzanian minerals to power mining communities. ##Problem Statement In Mwanza - Tanzania , Solar power output drops by 30% during the rainy season. This causes blackouts for communities and financial losses for mini-grid operators like JUMEME. Currently solar design tools do not factor in local weather patterns or locally available minerals. ## Solution An AI prototype that predicts solar panel efficiency based on: 1. **Mineral composition**: Hypothesis - Tanzanite+3% efficiency, High-Grade silica +2% 2. **Weather Forecast**: Rainy season reduces output 3. **Monthly Irradience**: Jun - Aug has peak sun in Mwanza - 'Tanzanite + Aug + Sunny = 23% efficiency' - 'Tanzanite + Apr + Rainy = 16% efficiency' - 'High-Grade silica + Apr + Rainy = 15% efficiency' - ## How to Run 1.Open "tanzanite_ai.ipynb" in google colab 2.Run all cells 3.Change mineral, month, weather parameters to test ## Research Basis This work is informed by: 1.UNECA. "Prospects of Africa's minerals for renewable energy technologies" 2.MDPI. " The role of transition metal oxide in perovskite solar cells" 3.arXiv. "Low-Cost perovskite solar cells from abudant materials" ## Data - Panel cost 550w Mono: Tsh 265000-285000 [source: Jiji Tz 2026] - Location: Mwanza TZ - Base Efficiency: 18% ## Built For Microsoft Research AI & IBM Research Africa Application 2026. Author: SELEMAN MAGANGA MICHAEL - Unshakable Energy AI ## License MIT License