|
Cycle 9 (2020 Deadline) Morogoro youth empowerment through establishment of social innovation (YEESI) lab for problem-centered training in machine vision PI: Kadeghe Fue (kadefue@sua.ac.tz), Sokoine University of Agriculture U.S. Partner: Glen Rains, University of Georgia Project Dates: May 2021 - February 2024
Project Webpage
Project Overview: The project proposes to establish a social innovation lab (YEESI) for a machine vision program that will be used by youth in the Morogoro region of Tanzania. There are young people in the area who have studied information technologies and allied sciences, and while most of them can write computer programs, they cannot solve machine vision problems. This project aims to increase awareness among the youth of Morogoro and nearby regions to address machine vision problems in agriculture. Machine vision is a new and understudied practice in Tanzania; hence, this project will contribute to efforts in the creation of scientific societies that address the most pressing problems faced by more than 80% of Tanzania’s population who engage directly in farming.The project expects to train more than 50 young technology enthusiasts who will be able to address the most pressing problems in agriculture and develop advanced digital tools to solve these problems. The main agricultural problems can be classified into five categories, as explained below:
- Disease Detection and Classification: The project will develop experts who will solve problems in disease identification using machine vision for most of the diseases in crops and livestock, which are misdiagnosed by farmers.
- Weed Classification: The project will develop algorithms that accurately identify weeds and contribute to the growing scientific database for automatic weed detection.
- Pest Detection and Classification: Appropriate tools using machine vision for Integrated Pest Management (IPM) are needed in Tanzania, as IPM has been hindered due to a lack of extension officers to train farmers on mitigation and identification of pests in agriculture.
- Crop Seedlings Stand Count and Yield Estimation: Use of machine vision and drones instead of scouting manually to estimate stand counts would provide appropriate mitigation strategies for replanting that would be beneficial to commercial farmers. Also of importance are algorithms to sort and estimate yield by counting the fruits and to estimate the amount of other agricultural products.
- Crop Vigor Estimation: Most farmers apply inputs evenly across the farm because they cannot predetermine crop vigor. Accurate estimation of crop health would help farmers to mitigate the problems earlier and improve crop performance and avoid failure. Algorithms to determine crop vigor developed in this project will contribute to the improvement of the methods to estimate crop performance earlier.
The proposed project is expected to have several development impacts. Technologies that are going to be developed by youth will be used for data collectors, data labelers, and systems developers who will be employed on a short-term or long term basis. Some will become innovators and entrepreneurs who can develop start-ups, spin-off, and innovative companies. Youth engaged in this project will also develop an interest in farming knowledge that would be crucial in the development of agriculture in the country and inspire other youth to engage in farming. Farmers who will use tools developed from this project will improve knowledge in crop management. These tools will help protect the environment, as they will enable farmers to produce prescriptive maps to help them to perform variable-rate application of pesticides and other farm inputs as determined by machine vision. The tools will also support farmers’ decisions on crop production by helping them avoid less fertile land and better control pests and diseases.
Final Summary of Project Activities
The YEESI Lab at Sokoine University of Agriculture is a center of excellence for advancing ICT education in Tanzania. Supported by PPER, The lab has been at the forefront of innovation and has designed problem-based and student-centered methodologies that have helped bridge the gap between theoretical knowledge and practical skills.
The lab's hands-on approach to learning has enabled it to host training workshops and engage students in competitions that provide essential skills needed in the tech industry. These skills include startup establishment, fundraising, and specialized training in Machine Learning (ML), Machine Vision (MV), and Natural Language Processing (NLP) applications. This approach has not only enhanced the technical proficiency of students but also fostered significant personal growth.
Furthermore, the lab has been committed to promoting gender equality in the tech industry by encouraging female students to participate in competitions and challenges, such as those presented by Zindi African Challenges and hackathons, making them more competitive in job markets. These competitions have helped showcase the capabilities of students on various scales, and they have gained recognition and awards that validate their skills and increase their visibility and credibility within the tech community.
One of the exceptional aspects of the YEESI Lab's efforts is its focus on technological advancement and community impact, which has facilitated substantial local development. By fostering innovation and building capacity, the lab has contributed to sustainable agricultural practices and economic growth within Tanzania. The lab has helped establish five Artificial Intelligence start-ups that develop technologies to assist farmers in making data-driven decisions. This integration of diverse disciplines has also cultivated an entrepreneurial mindset among students, positioning them well for future endeavors in technology and entrepreneurship.
The YEESI Lab has trained students and worked with start-ups and NGOs to establish the first comprehensive machine vision dataset suitable for central regions of Tanzania. The dataset has been open-sourced to allow more collaborators to work on it while developing Machine Learning models that would be useful in Tanzania and Morogoro region. The startups can use the data for free. The lab has also collected more data using drones, which involved training farmers on the beneficial use of drones, since it is a new technology to most remote areas where most farmers are located. The farmers were able to understand how drones work and present problems that they think drones may help solve, including issues on salinity and floods.
It is worth noting that the YEESI Lab's approach to problem-based learning has yielded exceptional results, and some achievements have been made by students who were not enrolled in ICT degree programs. This demonstrates the effectiveness of involving multiple disciplines in ICT-related problem-solving and innovation.
Through these comprehensive efforts, the YEESI Lab has enhanced the educational landscape in ICT and contributed significantly to community entrepreneurship, empowerment, and sustainable development. The lab has positioned itself as a leading center of excellence in ICT education in Tanzania, and its innovative methodologies have become a benchmark for other institutions seeking to improve their education systems.
Publications
Parab, C. U., Mwitta, C., Hayes, M., Schmidt, J. M., Riley, D., Fue, K., ... & Rains, G. C. (2022). Comparison of Single-Shot and Two-Shot Deep Neural Network Models for Whitefly Detection in IoT Web Application. AgriEngineering, 4(2), 507-522. https://doi.org/10.3390/agriengineering4020034
Kiobia, D. O., Mwitta, C. J., Fue, K. G., Schmidt, J. M., Riley, D. G., & Rains, G. C. (2023). A Review of Successes and Impeding Challenges of IoT-Based Insect Pest Detection Systems for Estimating Agroecosystem Health and Productivity of Cotton. Sensors, 23(8), 4127. https://doi.org/10.3390/s23084127
Zhang, J., Hu, Y., Li, F., Fue, K. G., & Yu, K. (2024). Meta-Analysis Assessing Potential of Drone Remote Sensing in Estimating Plant Traits Related to Nitrogen Use Efficiency. Remote Sensing, 16(5), 838. https://doi.org/10.3390/rs16050838
Back to PEER Cycle 9 Grant Recipients
|
|
|
|