Namibia’s voice on data: Co-designing a New and Open Dataset for Rangeland and Pasture Management.

Posted on April 30, 2024

Representatives of key stakeholders in rangeland management from the public and private sectors as well as researchers, students and farmers convened on the 23rd of April at Roof of Africa in Windhoek. The objective was to collaboratively design the upcoming open dataset for rangeland and pasture management for Namibia.

This innovative project aims to revolutionize Namibia’s livestock sector by providing essential data to formulate accurate predictions and help informed decision-making. For that, Farm4trade, a Namibian based start-up has joined hands with the University of Namibia, Omeva Consulting and local experts to create an open dataset pairing satellite images with collected ground information. The aim: to facilitate the development of tools for pasture management, including the calculation of carrying capacity and biomass.

Dr. Ercole, the CTO and founder of Farm4trade warmly welcomed the guests and explained the vision of Farm4trade of leveraging data for a more sustainable, efficient, and traceable livestock sector. Maria Luisa de la Puerta, CEO of Omeva Consulting and facilitator of the event presented the project to the attendants as well as illustrated the benefits and needs of data science in agriculture encouraging guests to actively participate.

Prof Simon Angombe, Associate Dean of the School of Agriculture and Fishery Sciences of the University of Namibia shed light on their several projects on rangeland management and the University’s vision on promoting sustainable rangeland management and utilization. To foster cooperation and information exchange, flagship projects of the University of Namibia were invited to present, such as the project “Rangeland Improvement through Bush Control and Sustainable Intensification to Mitigate Climate Change and Improve Livelihoods and Food Security in Southern Africa (Ribs)” or “Enhancing Community Adaptation through Climate Resilient Agriculture: Encroacher bushes value chains initiative in Otjozondjupa Region, Namibia”. A special mention deserves the engaging discussion sparked among the attendees by Mr. Hiskia Athingo's presentation on the RIBS project.

Interesting open talks about datasets were the highlight of the event. Attendees were divided into groups, participating in interactive exercises designed to elucidate the importance of having Findable, Accessible, Interoperable, and Reusable (FAIR) datasets. Groups were asked to search for datasets on google and were asked to describe their experience and provide inputs on how to improve them. Most of the attendants expressed their frustration in understanding some of the target datasets and provided inputs on how to improve them. “It would help to have a detailed methodology, including the data collection time, site and methods”- said Dr. Maria Shipandeni, lecturer and animal nutritionist from the University of Namibia. Dr. Kapuire, a lecturer at NUST added – “Appropriate descriptions and keywords can aid in locating data sets, as can open-source platforms”. Other suggestions included having a FAQ session, allowing to download specific sections of the dataset separately, and offering a usage sample. Everyone agreed that Kaggle was the easiest to find and navigate platform.

“Our dataset needs to be found and understood by a broad range of experts so that it shed the much-needed light on sustainable rangeland utilization. We decided to seek insights from end-users on how to accomplish this.” said Valerio Perucci, Farm4trade’s Project Manager.

After that, the groups were asked to come up with a use case for the dataset. This exercise enabled cooperation among stakeholders of different institutions, backgrounds, and ages.

One of the most anticipated sessions was the interactive panel discussion where attendees could ask questions and share their thoughts. Experts from different fields discussed the challenges and opportunities in rangeland management, emphasizing the importance of collaboration and continuous learning. The insights gained from these discussions will be invaluable in refining the dataset and ensuring it meets the needs of all stakeholders.

Another key highlight was the demonstration of new technologies in data collection and analysis. Participants were introduced to advanced tools and methods that can enhance the accuracy and efficiency of data gathering. This hands-on session was particularly beneficial for young researchers and students, who were eager to apply these techniques in their own projects.

The event also served as a platform for forging new partnerships and strengthening existing ones. Many attendees expressed their enthusiasm for future collaborations, recognizing the immense value that shared data and cooperative efforts can bring to the field of rangeland management. The collective commitment to advancing sustainable practices through innovative data solutions was palpable, leaving everyone motivated to continue the journey towards more informed and effective rangeland management strategies.

Prof. John Mupangwa, Professor and Animal Scientist at the University of Namibia, presented the audience's his group’s use case - “The dataset should be used at the national level, our use case highlights its potential to be used by governmental institutions to be used by farmers all over the country”. Dr Kapuire’s group on the other hand proposed potential applications for the data for underserved communities, such as Ovahimba and San people.

The event was also marked by the present role of the youth, with Mr. Deon Gewers, a final year NUST student, and Lapaka Petrus, a recent UNAM graduate, taking the lead in presenting the methodology. They elaborated on their tasks, challenges, and successes.

“Data science is a new and booming area full of possibilities, and it is our duty to empower and involve the youth to lead the way forward. Today they had a chance to prove their potential, and it has been a true pleasure to stand alongside them”- Ms Maria Luisa de la Puerta, Project manager and facilitator.

The event concluded with a call to action for all participants to stay engaged and proactive in their respective fields. Attendees were encouraged to continue sharing knowledge, resources, and experiences to further enhance the impact of the dataset. Plans were made for regular follow-up meetings and workshops to track progress and address any emerging challenges.

In conclusion, attendants agreed on the potential of datasets to help Namibian farmers, to make their farming more efficient, and to help mitigate climate change by and informed policy making. Access to the gathered data will be granted to institutions, Farmers Unions, and Associations, with the option to distribute it throughout their network. Academics and students would also gain from it, as it will empower students and promote research. The project has been funded by Lacuna Fund, an organization that aims at enabling labelled datasets for machine learning tools with high social value in low- and middle-income contexts worldwide, under the call "Labelled Agricultural Datasets for Machine Learning Solutions in Sub-Saharan Africa".

The dataset will be launched in the upcoming weeks. Interested stakeholders can reach out to info@farm4trade.com or maria@omevaconsulting.com for more information.

Join the Movement for Sustainable Rangeland Management!