The AI Lab at the German Environment Agency

The AI Lab, comprising approximately 30 staff members, utilizes AI and Big Data for environmental and sustainability applications. It serves as an innovation and experimentation space for the four federal agencies operating under the auspices of the  Federal Environment and Consumer Protection Ministry (BMUV) to demonstrate the added value of AI for both humanity and the environment, while addressing research questions related to the sustainable and responsible use of AI. The objective is to unlock AI as a key technology not only for industry but also for politics and research, particularly in the fields of environmental and climate protection.

The AI Lab develops applications that support the diverse tasks of the environmental sector—ranging from species protection to radiation protection, from nuclear safety to climate change adaptation and environmental monitoring. The initial implemented  examples of AI applications include:

• Identifying wind turbines in satellite data to support the planning of the energy transition,
• Detecting protected species on online trading platforms, and
• Monitoring hazardous materials and mitigating risks through the identification of radioactive nuclides via gamma spectrum analysis.

 

The external link opens the BMUV website, featuring a movie clip of the opening of the AI Lab in Dessau on October 13, 2023.

Movie clip of the AI Lab Opening (German)

 

To identify ideas and requirements for AI- and data-based applications, prepare them for processing by the AI Lab, and prioritize them for prototyping, the inter-agency process of „Use Case Discovery“ has been developed.

As a cyclical process, the Use Case Discovery structures the collaboration between the AI Lab and the specialized divisions of the authorities in the environmental sector. It ensures efficient resource utilization within the AI Lab and provides transparency and comparability among the various use cases.

The AI Lab's process of use case discovery.

Source: AI Lab at UBA

The use cases require close collaboration and intensive exchange between the AI Lab and the specialized departments of various authorities. The success of a project depends equally on both parties. The specialized departments benefit in various ways from this collaboration, such as through knowledge transfer, support in tendering processes, and the implementation of software. The created added value can thus be best illustrated through the perceptions of the specialized departments.

„Armed with interdisciplinary expertise, the AI Lab is helping us to finally bring a long-anticipated project to fruition.“

Scientific advisor, Federal Office for Radiation Protection

„The AI Lab develops scripts that enable me to better investigate the statistical relationships between air pollutants and meteorological variables. Given the numerous transformation and transport processes between the source of an air pollutant and its impact on people, I am very grateful for the methodological support and advice.“

Technical employee, German Environment Agency

„The valuable insights from the cooperative workshops of the Lab are already actively supporting our internal project development processes, even beyond our joint use cases. We look forward to the upcoming first prototypes and further collaboration.”

Head of department, Federal Agency for Nature Conservation (BfN)

„Significant time savings! Relevant AI developments for the BfN can rapidly be prototypically developed and tested for usability. This reduces the need for extensive research projects with long durations and initiation phases in the scope of AI topics.“

Digital strategist, Federal Agency for Nature Conservation (BfN)

The AI Lab is committed to responsible, values-based software development that takes into account both human and environmental factors, as well as potential interactions. In this context, not only technological feasibility is considered, but also the potential impacts of a development project in regard to socio-ecological aspects. All use cases are analyzed in this respect, with the understanding of ethics as a process rather than a checklist; that is, applicable ethics are integrated into daily work.

Developing responsibly

The AI Lab places particular emphasis on the responsible handling of data and develops solutions for the resource-efficient use of AI and Big Data (Responsible & Green AI). Various aspects of sustainable software development are addressed, including the energy-efficient use of hardware, the careful and ethical selection of data and algorithms, the improvement of energy efficiency in existing AI models, and the usability by third parties in the context of open source.

Source: Paul Skorupskas / unsplash.com/de/@pawelskor

The AI Lab is an initiative within the Digital Policy Agenda for the Environment of the BMUV and is part of the BMUV's 5-Point Program "Artificial Intelligence for Environment and Climate" (German). For this purpose, €26.4 million is available from the federal government's Economic Stimulus and Future Package (2021). Approximately 30 staff members will be employed, initially on a temporary basis until 2025, at the locations of Leipzig, Berlin, and Dessau-Roßlau. The interdisciplinary team includes experts from various fields, such as project management, data science, data engineering, high-performance computing, AI ethics, remote sensing, user experience, and interface design.

The AI Lab is defined as a measure in the update of the AI Strategy 2020 (German), the Data Strategy of the Federal German Government 2021, and as a milestone in the German Recovery and Resilience Plan (DARP).