A Cyber-Physical Architecture for Crop Recommendation: Integrating Climate Prediction (XGBoost), IoT Sensing, and Multi-Criteria Decision-Making

• Vitor Hugo Barbosa Melo UFPA
• André Vinicius Neves Alves UFPA
• Saulo Mattheus Ribeiro de Oliveira UFPA
• Waldemiro José Assis Gomes Negreiros UFPA
• Filipe Corrêa da Silva UFPA
• Nilton Rodolfo Nascimento Melo Rodrigues UFPA
• Diego Lisboa Cardoso UFPA
• Marcos Cesar da Rocha Seruffo UFPA

DOI: https://doi.org/10.5753/wcama.2026.23738

Resumo

This work proposes a unified cyber-physical framework to evaluate the agronomic suitability of Açaí, Cocoa, and Soybean in the state of Pará, Brazil. The methodology integrates ERA5 climate data, XGBoost forecasting, and a conceptual IoT node for local microclimatic calibration using Edge Computing. Climate projections for 2026–2028 fed a modified SAW (Simple Additive Weighting) decision engine. For Soybean, the system identifies optimal planting windows (January–April), while for Cocoa and Açaíit supports year-round thermal and water stress monitoring. Simulations indicated high suitability during the rainy season and severe penalties in dry months (July–October), highlighting Cocoa’s vulnerability to heat stress and Soybean’s planting restrictions. Results validate the software layer and support future IoT field deployment.

Referências

Abreu, A. L. and Barbosa, I. C. (2021). Desenvolvimento da soja para determinação do período de plantio em tomé-açu/pa. Master’s thesis, UFRA, Tomé-Açu, PA.

Chakri, A. et al. (2025). Advanced forecasting using era5 and machine learning in tropical regions. Agricultural and Forest Meteorology, 320:108900.

FAPESPA (2024). Relatório do PIB do pará – 2023. Technical report, Fundação Amazônia de Amparo a Estudos e Pesquisas, Belém, PA.

Fishburn, P. C. (1967). Additive utilities with incomplete product sets. Operations Research, 15(3):537–542.

Garg, R. et al. (2021). Iot-based precision agriculture: A comprehensive review. IEEE Internet of Things Journal, 8(15):11850–11875.

Gong, L. et al. (2025). Edge computing architectures for real-time agricultural monitoring. IEEE Transactions on AgriFood Electronics, 3(2):112–125.

Ministério da Agricultura e Pecuária (2024). Portaria spa/mapa nº 5/2024: Zarc para a cultura do açaíirrigado. Diário Oficial da União, Seção 1, p. 7. Brasília, DF.

Nangoi, I. I., Luiz, A. J. B., and Sentelhas, P. C. (2006). Determinação dos parâmetros de risco climático para a cultura do cacau. ESALQ/USP.

Nawaz, M. et al. (2025). Iot and ai for smart agriculture in resource-constrained environments. Discover Agriculture, 2(1):15.

Oses, N. et al. (2020). Machine learning applications for agricultural planning using era5 reanalysis. Computers and Electronics in Agriculture, 175:105500.

Serra, R. and Sodré, G. A. (2021). Manual do cacauicultor: perguntas e respostas. Technical Report Boletim Técnico 221, Embrapa.