The OECD.AI Policy Navigator

Our policy navigator is a living repository from more than 80 jurisdictions and organisations. Use the filters to browse initiatives and find what you are looking for.

AI Brownfield Site


Added by:   OECD analyst
Added on:   23 Sep 2026
Updated by:   OECD analyst
Updated on:   23 Sep 2026

Austria has great potential in terms of industrial and commercial brownfield sites. The initiative is a nationwide public AI-based map that identifies these brownfields primarily based on geodata and aerial photographs. It was created to provide reliable information on unused industrial and commercial sites for planners, municipalities, developers and researchers. The goal is to support land recycling and reduce land consumption by making these opportunities visible.

Name in original language

KI Potenzialflächenkarte

Initiative overview

The initiative addresses Austria’s persistent challenge of high land consumption, around 6.5 hectares per day, far exceeding national sustainability targets. A major opportunity lies in reactivating brownfields: previously industrial or commercial sites that are unused or underused. Earlier studies estimated up to 10,000 such sites nationwide, yet no comprehensive, publicly accessible dataset existed.

Against this backdrop, the initiative seeks to leverage AI-based image analysis to systematically identify brownfields across Austria and support more sustainable land use, prioritising redevelopment over new land conversion. The project’s core objective is to create the first nationwide, data-driven brownfield map using optical remote-sensing data. By applying a trained AI classification system to orthophotos, approximately 65,000 industrial and commercial sites larger than 1,000 m² were analysed. About 4,700 brownfields, sites with very high prediction confidence, covering roughly 4,000 hectares were published on the public WebGIS platform and made available as open government data.

The initiative aims to support municipalities, planners, investors, and policymakers in prioritising land recycling, aligning with national commitments to reduce soil sealing and promote redevelopment over new zoning. By making the data publicly accessible through www.brachflaechen-dialog.at and data.gv.at, the project institutionalises a standardised, repeatable method for monitoring brownfield potential.

Looking ahead, the initiative is designed for continuous evolution and long-term integration into Austria’s land-use governance. Since one-third of the country is re-photographed annually, updated orthophotos will feed into the system, enabling regular refreshes of the brownfield map. This creates a dynamic monitoring tool rather than a one-time inventory. Through institutionalisation within the Federal Ministry and ongoing technical updates, the initiative is positioned to become a permanent, scalable instrument for reducing land consumption and promoting sustainable spatial development.

Other relevant details

Results, outcomes and impacts: Since its release in December 2025, the brownfield map has already generated strong interest from real-estate developers and research institutions, demonstrating its immediate relevance. Early results include the identification of 4,700 brownfields covering about 4,000 ha. Impacts so far are measured through AI-based orthophoto classification and validation with regional data. While concrete effects on land recycling cannot yet be measured, the high demand suggests that the dataset will support future redevelopment decisions and contribute to reducing land consumption. Challenges: Key challenges included ensuring AI accuracy with limited training data and relying on high-resolution aerial images that are only updated every three years. This limits how quickly changes in brownfield status can be detected. The team addressed this through close cooperation with regional agencies and iterative model improvements. Lessons learned: A key lesson is that high-quality reference data and stable institutional partnerships are essential. Future success will depend on integrating lower-resolution but frequently updated satellite imagery, enabling more regular detection of emerging or reactivated brownfields and ensuring the long-term relevance of the monitoring system.

About the policy initiative


Category:

  • AI policy initiatives, programmes and projects

Initiative type:

  • AI use cases/projects in the public sector

Status:

  • Active

Start Year:

  • 2024

OECD AI Principles: