Insects and pests monitoring is crucial for commercial crops to comply with modern phytosanitary requirements. Some countries may require information on controlled pests and insects. This device offers an automated solution through imaging yellow traps and using machine learning algorithms to count the pests.

This project is endorsed by FAO

In this project, we are collaborating with HYDROC GmbH, a German-based company, to integrate Delft-FEWS with web mapping techniques for online flood model visualization. Delft-FEWS is a state-of-the-art flood forecasting and early warning system that can simulate and track rainfall, predicting floods.

This project is directly funded by EU.

This is an innovative project led by the Lab, with support from SUA, TWAS, BMBF, and UNESCO. Our proposal revolves around the development of an autonomous robot capable of safely and accurately applying pesticides and other chemicals to crops. Our primary aim is to create an affordable unmanned ground vehicle (UGV) that utilizes a Center-articulated Oscillating Chassis Robot design for variable-rate pesticide application. The key objective of this endeavour is to design a centre-articulated and oscillating chassis robot that can independently carry out tasks such as crop spraying, collecting phenotypic data (such as height and canopy size), plant counting, as well as pest and disease detection. The specific objective of the project is

i)    To fabricate the prototype of the AgroSpray robotic system

ii)   To develop autonomous navigation algorithms and software for the AgroSpray system

iii)  To integrate units for an autonomous spraying robot (AgroSpray)

iv)  To evaluate the performance of the AgroSpray system at the SUA Model Farm

v)   To develop a production line and establish a business model for the AgroSpray system

Principal Investigator of the Project: Dr. Kadeghe Fue, Sokoine University of Agriculture

Project Dates: December 2022 - December 2024

It is important to note that precise prediction of crop yields on a farm level can be of great value to smallholder farmers, as it allows them to estimate their net profit. Additionally, this information enables insurance companies to determine appropriate payouts and facilitates agricultural-related loans for farmers. 

The key project objectives include:

• To develop a model that utilizes historical multi-source data to predict maize and sorghum yield at the district level 

• To deploy a small-scale smart farming system using low-cost Internet of Agro Things (IoAT) sensors and interactive cloud-based big data analytics to monitor and evaluate crops’ performance in real-time.

• To pilot a big data model to predict farm-level yield using low-cost agricultural Internet of Agro-Things (IoAT) sensor data by enhancing district-level resolution yield data to farm-level resolution yield data using Generative Adversarial Networks (GANs).

• To conduct the economic feasibility of using agricultural IoT and big data for small-scale farm monitoring and yield prediction.

• To formulate a data-driven policy brief on crop prediction using multi-source big data. The research will identify and reach potential farmers to use agricultural IoAT.

Team Members:

1. Dr. Alcardo Alex Barakabitze — Principal Investigator (PI)

2. Prof. Camilius A. Sanga — Co-PI

3. Sahara Ventures — Co-PI (collaborator)

4. RECODA — Co-PI (collaborator)

5. Dr. Kadeghe G. Fue — Team member

6. Dr. Neema Nicodemus — Team member

7. Dr. Michael P. J. Mahenge —  Team member

8. Dr. Joseph P. Telemala — Team member

Project Dates: December 2022 - June 2024

Website: https://ai4morecrops.co.tz/wp/ 

The primary objective of this project is to gather data related to three waterborne diseases prevalent in Tanzania: typhoid fever, diarrhea (particularly dysentery), and amoebiasis. The collected data will serve as a valuable resource for research and the implementation of predictive machine learning (ML) models in the field of medicine and healthcare within the region. To ensure the dataset is comprehensive and suitable for predictive ML purposes, existing scattered data available in printed documents will be aggregated and supplemented with additional sources. The project aims to collect primarily historical data over a span of five years, from 2017 to 2022, focusing on four Tanzanian regions: Morogoro Municipal Council (MC), Singida Municipal Council (MC), Dodoma City Council (CC), and Dar es Salaam City Council (Temeke MC, Ilala MC). These regions have been selected due to the high prevalence of typhoid fever, diarrhea (including dysentery), and amoebiasis. While existing data within the healthcare sector of Tanzania, such as hospital records and other healthcare facilities in the study area, will be considered, efforts will be made to digitize and structure this data for ML applications. Additionally, supplementary data on water sources and weather conditions will be collected from online sources to complement the dataset.

Principal Investigator of the Project: Dr. Neema N. Lyimo, Sokoine University of Agriculture

co-PI: Dr Joseph P. Telemala

  Dr Silvia Materu

  Dr Ndimile C. Kilatu

  Dr Kadeghe G. Fue

Project Dates: January 2023 - May 2024

Our project, titled "Prediction of the Rain-fed Maize Yield in Tanzania using Time-series Machine Learning Models (TanPredikt)," is receiving support from DSA in collaboration with Deep Learning Indaba. This support is made possible through the Artificial Intelligence for Development in Africa Program (AI4D) Research Grant provided by the International Development Research Centre (IDRC).

In agriculture, various parameters can be utilized to train algorithms for predicting crop performance. These parameters encompass crop-related information such as growth stages and leaf area index, soil information including soil type, soil pH, and soil cation exchange capacity, environmental factors like rainfall and precipitation, nutrient levels encompassing natural and added nutrients, field management practices including irrigation, fertilization, and chemical application, solar information such as gamma radiation, temperature, photoperiod, and shortwave radiation, vegetation indices like normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI), as well as wind speed, pressure, historical yield information, and images.

Our project aims to leverage the most advanced time-series machine learning algorithms, utilizing all available datasets at our disposal. By doing so, we anticipate a significant improvement in the prediction of maize crop performance, surpassing the limitations of existing methods like the Water Requirement Specification Index (WRSI), which relies predominantly on water-based information and offers limited insights.

Principal Investigator of the Project: Dr. Kadeghe Fue, Sokoine University of Agriculture

Project Dates: December 2022 - November 2023

Towards a Nationwide Automatic Irrigation Scheduling System using Geospatial Artificial Intelligence (NAISS-GAI), a SUARIS Grant

Principal Investigator of the Project: Dr. Kadeghe Fue, Sokoine University of Agriculture

Project Dates: Feb 2022 - Jan 2024

Collaborators: Prof. B. Mbilinyi, Prof. C. Sanga, and Dr. W. Mbungu

website: https://www.sua.ac.tz/announcements/list-successful-projects-under-sua-research-and-innovation-support-suaris-2nd-phase

Morogoro YEESI Lab is a PEER Project hosted at the Sokoine University of Agriculture. This project is funded by National Academy of Sciences, US Agency for International Development and US Department of Agriculture. This project received PEER Cycle 9 Grant.

Principal Investigator of the Project: Dr. Kadeghe Fue, Sokoine University of Agriculture

U.S. Partner: Prof. Glen Rains, University of Georgia

Project Dates: May 2021 - Feb 2024

Collaborators: Prof. Camilius Sanga (SUA), Dr. Alcardo Barakabitze (SUA), Dr. Wulystan Mtega (SUA), Prof. Siza Tumbo (Regional Administrative Secretary, Shinyanga) and Prof. Madundo Mtambo (Director General, Tanzania Industrial Research and Development Organization )

website: www.yeesi.org

This project will phase out in February 2024, the website www.yeesi.org will be hosted in EPAL as www.yeesi.epalab.org 

With the support of the German Federal Ministry of Education and Research (BMBF), TWAS launches a new programme to strengthen the capacity of African countries lagging in science and technology. The new programme is aimed at young scientists who are getting established in their country or about to return home to an academic position. Under this scheme, grants are awarded to promising high-level research projects in Agriculture, Biology, Chemistry, Earth Sciences, Engineering, Information Computer Technology, Mathematics, Medical Sciences and Physics carried out in African countries lagging in science and technology identified by TWAS.

Principal Investigator of the Project: Dr. Kadeghe Fue, Sokoine University of Agriculture

Project Dates: Jan 2022 - Dec 2023

Website: https://twas.org/opportunity/seed-grant-new-african-principal-investigators-sg-napi

FSNet-Africa aims to design and implement food systems research in partnership with stakeholders to identify solutions that can bring about sustainable change in African food systems.

In this project, we are implementing The smartphone Decision Support for Variable Rate Fertilizer Application (smartphone - VRA )

Principal Investigator of the Project: Dr. Kadeghe Fue, Sokoine University of Agriculture

Project Dates: July 2021 - Dec 2023

Collaborators: DPRTC Office

Website: https://fsnetafrica.com/fellowship/our-fellows/

The project will identify novel genes or QTLs involved in flood or salinity tolerance of rice using African rice germplasm such as Oryza glaberrima and wild relatives. Three underlying knowledge gaps for achieving the overall objective are i) the role of the root barrier to radial O2 loss (ROL) in protecting against tissue intrusion of soil phytotoxins and salt (NaCl), ii) the genotypes and role of superhydrophobic leaf cuticles resulting in formation of leaf gas films during submergence and the consequences for sustaining gas exchange and protecting against salt intrusion and iii) the prevalence of the trait of anaerobic germination. We propose to use promising genotypes of wild relatives from wetland habitats to uncover these trait capacities 

Principal Investigator of the Project: Prof. Susan Msolla, Sokoine University of Agriculture

Project Dates: 2021 - March 2025

Collaborators: Dr Kadeghe Fue (I am co-supervising a PhD student on Digital DSS for Salinity Management) 

Website: ​​​​​​​https://climate-smart-rice.com/