GIZ Data Lab...

Unconventional, Innovative, Experimental - The GIZ Data Lab is a platform that brings together thinkers and practitioners to promote the effective, fair and responsible use of digital data for sustainable development. Since its foundation in January 2019, the Lab applies an agile and experimental way working to explore new trends and develop forward-looking solutions in GIZ partner countries. Together with a strong network of partners, the team promotes the use of digital data to advance development and to support the delivery of innovative services by GIZ. The Data Lab team conducts various experiments in the field of data for development. Experiments are limited in time and scope and seek to test a concrete hypothesis. By working with a broad range of partners, the GIZ Data Lab aims at building a strong network and enhance learning within the organization. All in all, the team strives to be at the forefront in applying innovation and the latest standard, enabling them to offer the best solutions in an ever-evolving market in the field of development and beyond.

Great potential of non-traditional data for real impact 

In an increasingly digitalized world, the use of non-traditional data sources, such as satellite or mobile data, has great potential for improving our development interventions. At the same time, however, increased use of novel data sources and analytical techniques also raises crucial ethical and privacy issues.  

The multifaceted experiments of the GIZ Data Lab serve to develop and test data-based methods in development cooperation, always against the background of the described continuum between opportunities they hold for sustainable impact and challenges that mitigate this process. 

Strong partnerships and effective learning 

A strong focus on partnerships – in innovation, implementation and advocacy - is of great importance for the GIZ Data Lab. As an integral part of its work the team engages with some of the most renowned actors in the data innovation space to co-design and implement initiatives, benefit from each other's strengths and thus, can have a bigger impact than each of us would be able to achieve individually.  

Learning and Working out Loud are key principles that the Data Lab is driving forward to increase its impact. The main channels through which the team shares its work with others are the GIZ Data Lab Blog and Twitter Channel.   

Get in touch if you want to collaborate and learn more about our work.  

 

Below you find some of the experiments we are working on.  

Experiments

The use of Artificial Intelligence (AI) in agriculture abounds, varying from pest infection control to the employment of autonomous machinery, soil and crops health monitoring etc. Within the experiment “Artificial Intelligence for Agriculture”, the GIZ Data Lab explored the question: How is it possible to simplify the process of classification and yield prediction of cashew crops in Burkina Faso? Over the project the team employed public satellite imagery with GIZ-collected project data to develop an innovative data-driven crop information system for improved decision making. 

Correlating this data, the algorithm “learned” how to differentiate between cashew and non-cashew areas just by “looking” at the satellite images. This application showcases how development cooperation projects may unlock a “hidden data treasure” by designing their data collection procedures in compliance with AI training data requirements - a finding with great potential for supporting future data-based projects in the agricultural development sector. More information is available here.  

The use of the new data sources and analysis methods emerging in the context of ongoing digitalization and technological development can enable a more inclusive and sustainable form of (urban) transport planning. The Data 4 Mobility experiments in Dar es Salaam, Tanzania and Bangkok, Thailand aim to investigate whether it is possible to gain a better understanding of urban mobility by intelligently combining and evaluating different data sources such as satellite data, data from ride-hailing services etc. and thus contribute to a more sustainable form of transport and urban planning. 

In Dar es Salaam, a prototype dashboard was created, which connects various data sources. The dashboard illustrates the potential of intersecting relevant data sources and can help mobility planners to get a more accurate picture of their city. Find out more here.  

In Bangkok, prototypical models based on the ride-hailing data were trained, illustrating the potential of this new big data source in various relevant areas such as traffic flow analysis, speed profiling for certain route sections, volume forecasts and air quality estimates. Read more about the experiment here.  

Women and girls are among those that are most affected by Covid-19, for example due to economic or social constrains. However, also more invisible consequences that have now begun to be uncovered. In this experimental project, the GIZ Data Lab focuses on one of these more indirect and invisible consequences of the Covid-19 pandemic, which is causing alarm among many governments and human rights organizations worldwide: The reports of drastic increases in levels of gender-based violence, particularly violence experienced by women and girls. 

To optimize the response to what has been referred to as the “shadow pandemic”, this project seeks to unveil if a combination of big data and analytics can help better identify those at risk and channel support more efficiently to mitigate gender violence. The goal of this project is to develop an innovative approach to identify areas in Mexico in which the population could be at most risk of experiencing IPV, as well to generate key insights about the factors that are at play when IPV takes place, during and after a pandemic such as COVID-19. Therefore, the GIZ Data Labs aims to build a statistical IPV risk model that will translate into a map to visualize areas where women across the country are at most risk. 

For successful and sustainable project governance, it is essential to integrate the prototype and the associated learning experiences into more general subject areas. Accordingly, a dedicated Council for the Orientation of Development and Ethics (CODE) will be set up for the project. The CODE will also ensure that the cultural context and Do-No-Harm Principle are appropriately taken into the account throughout the further development of the project (i.e. incorporation of local masculinity concepts, aspects of recent refugee/migration movements, etc.). 

Historically, the Positive Deviance development approach was first applied in the 90s by Monique and Jerry Sternin in the field of child nutrition. Based on the observation that in every population there are individuals or communities who, despite facing similar challenges and limitations, achieve better results than their peers, this approach focuses on these outliers (or positive deviants) in order to discover unusual practices and strategies that solve successfully complex problems ­– particularly where conventional solutions failed. 

In an open learning and testing network, we partner with organizations such as UNDP Accelerator Labs or the University of Manchester and explore the use of new digital data sources in the systematic identification and understanding of positive outliers in various domains. 

Within the Data Powered Positive Deviance initiative the GIZ Data Lab and its partners initiated a total of 7 pilot experiments that test, implement and develop this approach in their respective context – ranging from finding positive deviants in rice-farming villages in Indonesia, safe public spaces for women in Mexico to districts that performed unexpectedly well during the Covid-19 Pandemic in Germany.