AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations

Smallholder agriculture faces a persistent paradox: while agronomic science and data availability have advanced rapidly, farm-level decisions are still guided largely by blanket recommendations and fragmented advice. Many digital tools now claim to offer site-specific guidance, yet most remain focus...

Descripción completa

Detalles Bibliográficos
Autores principales: Abera, Wuletawu, Mkuhlani, Siyabusa, Tibebe, Degefie, Assefa, Feben, Leroux, Louise, Seid, Jemal, Srivastava, Amit, Aguilar Ariza, Andres, Devkota Wasti, Mina, Ibrahim, Ali, Kouadio, Louis, Liben, Feyera, Llanos Herrera, Lizeth, Cherenet, Meklit, Corbeels, Marc, Kihara, Job, Vanlauwe, Bernard
Formato: Informe técnico
Lenguaje:Inglés
Publicado: 2025
Materias:
Acceso en línea:https://hdl.handle.net/10568/179828
_version_ 1855535525217173504
author Abera, Wuletawu
Mkuhlani, Siyabusa
Tibebe, Degefie
Assefa, Feben
Leroux, Louise
Seid, Jemal
Srivastava, Amit
Aguilar Ariza, Andres
Devkota Wasti, Mina
Ibrahim, Ali
Kouadio, Louis
Liben, Feyera
Llanos Herrera, Lizeth
Cherenet, Meklit
Corbeels, Marc
Kihara, Job
Vanlauwe, Bernard
author_browse Abera, Wuletawu
Aguilar Ariza, Andres
Assefa, Feben
Cherenet, Meklit
Corbeels, Marc
Devkota Wasti, Mina
Ibrahim, Ali
Kihara, Job
Kouadio, Louis
Leroux, Louise
Liben, Feyera
Llanos Herrera, Lizeth
Mkuhlani, Siyabusa
Seid, Jemal
Srivastava, Amit
Tibebe, Degefie
Vanlauwe, Bernard
author_facet Abera, Wuletawu
Mkuhlani, Siyabusa
Tibebe, Degefie
Assefa, Feben
Leroux, Louise
Seid, Jemal
Srivastava, Amit
Aguilar Ariza, Andres
Devkota Wasti, Mina
Ibrahim, Ali
Kouadio, Louis
Liben, Feyera
Llanos Herrera, Lizeth
Cherenet, Meklit
Corbeels, Marc
Kihara, Job
Vanlauwe, Bernard
author_sort Abera, Wuletawu
collection Repository of Agricultural Research Outputs (CGSpace)
description Smallholder agriculture faces a persistent paradox: while agronomic science and data availability have advanced rapidly, farm-level decisions are still guided largely by blanket recommendations and fragmented advice. Many digital tools now claim to offer site-specific guidance, yet most remain focused on prediction rather than decision making - estimating yields, risks, or responses without translating them into coherent, feasible choices for real farms. The result is a widening gap between scientific knowledge, digital intelligence, and the complex realities of smallholder farming systems shaped by soil constraints, climate variability, resource limitations, and household priorities. AgWise responds to this challenge by rethinking how agronomic intelligence is structured and deployed. Rather than treating artificial intelligence as a stand-alone solution, AgWise is designed as an embedded decision intelligence system that integrates agronomic science, data, models, expert knowledge, and AI within a single, coherent framework. Its core premise is that agronomic recommendations must be biophysically grounded, constraint-aware, and context-sensitive to be actionable at scale. AgWise therefore combines process-based crop and soil models, empirical evidence from field experiments, expert-driven rules, and AI-enabled orchestration to generate recommendations that respect both environmental realities and farmer objectives. Crucially, AgWise is modular, not monolithic. Different analytical engines serve distinct roles - models ensure physical consistency; data-driven methods capture observed responses, rules enforce feasibility and safety, and AI supports integration, explanation, and learning. This modularity allows AgWise to operate across diverse crops, agroecologies, and data environments, while remaining transparent, adaptable, and scientifically auditable. AgWise is not a chatbot, a black-box AI recommender, a single agronomic model, or a mobile application. AgWise is a modular decision-support architecture that converts inputs (data + models + rules + constraints + objectives) into a set of feasible advisory packages. By moving beyond isolated predictions toward integrated decision support, AgWise enables farmers, advisory services, and institutions to navigate real trade-offs—between productivity and risk, short-term gains and soil restoration, and profitability and resilience. In doing so, it offers a scalable pathway for translating decades of agronomic research into actionable, context-aware intelligence, supporting the transformation of smallholder agriculture toward greater sustainability, resilience, and impact. This document is a conceptual paper that articulates the design philosophy, decision logic, and system-level thinking underpinning AgWise. Its primary goal is to clarify what AgWise is, and to explain how agronomic advisories are structured within the system. The document is not intended as a technical manual or implementation guide. Detailed algorithms, parameterizations, and workflows will be described in complementary technical documents and scientific publications.
format Informe técnico
id CGSpace179828
institution CGIAR Consortium
language Inglés
publishDate 2025
publishDateRange 2025
publishDateSort 2025
record_format dspace
spelling CGSpace1798282026-01-15T02:15:42Z AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations Abera, Wuletawu Mkuhlani, Siyabusa Tibebe, Degefie Assefa, Feben Leroux, Louise Seid, Jemal Srivastava, Amit Aguilar Ariza, Andres Devkota Wasti, Mina Ibrahim, Ali Kouadio, Louis Liben, Feyera Llanos Herrera, Lizeth Cherenet, Meklit Corbeels, Marc Kihara, Job Vanlauwe, Bernard agronomic practices soil quality diversification fertilizer application digital innovation Smallholder agriculture faces a persistent paradox: while agronomic science and data availability have advanced rapidly, farm-level decisions are still guided largely by blanket recommendations and fragmented advice. Many digital tools now claim to offer site-specific guidance, yet most remain focused on prediction rather than decision making - estimating yields, risks, or responses without translating them into coherent, feasible choices for real farms. The result is a widening gap between scientific knowledge, digital intelligence, and the complex realities of smallholder farming systems shaped by soil constraints, climate variability, resource limitations, and household priorities. AgWise responds to this challenge by rethinking how agronomic intelligence is structured and deployed. Rather than treating artificial intelligence as a stand-alone solution, AgWise is designed as an embedded decision intelligence system that integrates agronomic science, data, models, expert knowledge, and AI within a single, coherent framework. Its core premise is that agronomic recommendations must be biophysically grounded, constraint-aware, and context-sensitive to be actionable at scale. AgWise therefore combines process-based crop and soil models, empirical evidence from field experiments, expert-driven rules, and AI-enabled orchestration to generate recommendations that respect both environmental realities and farmer objectives. Crucially, AgWise is modular, not monolithic. Different analytical engines serve distinct roles - models ensure physical consistency; data-driven methods capture observed responses, rules enforce feasibility and safety, and AI supports integration, explanation, and learning. This modularity allows AgWise to operate across diverse crops, agroecologies, and data environments, while remaining transparent, adaptable, and scientifically auditable. AgWise is not a chatbot, a black-box AI recommender, a single agronomic model, or a mobile application. AgWise is a modular decision-support architecture that converts inputs (data + models + rules + constraints + objectives) into a set of feasible advisory packages. By moving beyond isolated predictions toward integrated decision support, AgWise enables farmers, advisory services, and institutions to navigate real trade-offs—between productivity and risk, short-term gains and soil restoration, and profitability and resilience. In doing so, it offers a scalable pathway for translating decades of agronomic research into actionable, context-aware intelligence, supporting the transformation of smallholder agriculture toward greater sustainability, resilience, and impact. This document is a conceptual paper that articulates the design philosophy, decision logic, and system-level thinking underpinning AgWise. Its primary goal is to clarify what AgWise is, and to explain how agronomic advisories are structured within the system. The document is not intended as a technical manual or implementation guide. Detailed algorithms, parameterizations, and workflows will be described in complementary technical documents and scientific publications. 2025-12-23 2026-01-14T13:54:29Z 2026-01-14T13:54:29Z Report https://hdl.handle.net/10568/179828 en Open Access application/pdf Abera, W.; Mkuhlani, S.; Tibebe, D.; Assefa, F.; Leroux, L.; Seid, J.; Srivastava, A.; Aguilar Ariza, A.; Devkota Wasti, M.; Ibrahim, A.; Kouadio, L.; Liben, F.; Llanos Herrera, L.; Cherenet, M.; Corbeels, M.; Kihara, J.; Vanlauwe, B. (2025) AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations. CGIAR Sustainable Farming Science Program. 47 p.
spellingShingle agronomic practices
soil quality
diversification
fertilizer application
digital innovation
Abera, Wuletawu
Mkuhlani, Siyabusa
Tibebe, Degefie
Assefa, Feben
Leroux, Louise
Seid, Jemal
Srivastava, Amit
Aguilar Ariza, Andres
Devkota Wasti, Mina
Ibrahim, Ali
Kouadio, Louis
Liben, Feyera
Llanos Herrera, Lizeth
Cherenet, Meklit
Corbeels, Marc
Kihara, Job
Vanlauwe, Bernard
AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations
title AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations
title_full AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations
title_fullStr AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations
title_full_unstemmed AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations
title_short AgWise: Modular decision support tools integrating data, models, and AI for site-specific agronomy and soil health management recommendations
title_sort agwise modular decision support tools integrating data models and ai for site specific agronomy and soil health management recommendations
topic agronomic practices
soil quality
diversification
fertilizer application
digital innovation
url https://hdl.handle.net/10568/179828
work_keys_str_mv AT aberawuletawu agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT mkuhlanisiyabusa agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT tibebedegefie agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT assefafeben agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT lerouxlouise agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT seidjemal agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT srivastavaamit agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT aguilararizaandres agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT devkotawastimina agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT ibrahimali agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT kouadiolouis agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT libenfeyera agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT llanosherreralizeth agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT cherenetmeklit agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT corbeelsmarc agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT kiharajob agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations
AT vanlauwebernard agwisemodulardecisionsupporttoolsintegratingdatamodelsandaiforsitespecificagronomyandsoilhealthmanagementrecommendations