Objectives of the Product

Analysing Earth Observation (EO) data is complex and often inaccessible to non-technical users, involving multiple technical steps like dataset selection, preprocessing, and analysis. This complexity challenges organisations such as municipalities and environmental groups, which need data-driven decisions for issues like climate adaptation and infrastructure planning.

Despite the availability of EO data, there’s a gap between its potential and actual use. Large Language Models (LLMs) combined with geospatial processing can bridge this gap, but current solutions rely on proprietary technology. Actinia-copilot URBAN offers an open-source alternative, ensuring digital sovereignty and user protection.

Initially focusing on urban climate adaptation, it helps municipalities and planners gain actionable insights for initiatives like urban greening. The system allows users to analyse urban structures for climate adaptation using natural language interactions, reducing the need for geospatial expertise. Users can select areas or rely on the system’s suggestions based on indicators like sealed surfaces.

Actinia-copilot URBAN automatically acquires and processes relevant EO data, presenting results through interactive maps and reports to support decision-making. Actinia-copilot URBAN integrates open-source LLMs with the geoprocessing engine actinia, enabling natural language queries for EO data analytics. This shifts the complexity to AI, empowering users across sectors to extract insights easily.

Customers and their Needs

The targeted customers and users of the product are primarily public sector organisations, civil engineering and consulting firms, environmental monitoring organisations, and non-governmental organisations (NGOs). These stakeholders are actively involved in spatial planning, infrastructure development, environmental assessment, and climate adaptation activities, all of which require reliable, spatially explicit, and evidence-based information.

The primary user needs include easy access to high-quality EO data, transparent and reproducible analytical workflows, and actionable outputs that can be readily interpreted by non-specialists. Key challenges in meeting these needs are the technical complexity of EO data processing, limited in-house geospatial expertise, high costs of proprietary solutions, and concerns regarding data governance and digital sovereignty. The product addresses these challenges by providing an open-source, natural language-driven interface that automates EO data analysis.

Targeted customer/users countries

The targeted customers and users are currently primarily located in European countries with a strong potential for subsequent adoption across additional international markets.

Product description

The product called actinia-copilot URBAN, is an open-source, AI-enabled Earth Observation analytics system designed to support urban planning and climate adaptation. It integrates large language models with the actinia geoprocessing engine to enable users to perform complex EO data analyses through natural language interactions.

The system automatically translates user queries presented in natural language into reproducible geospatial workflows, handling data discovery, acquisition, preprocessing, and analysis across heterogeneous EO datasets. The core innovation lies in combining open-source large language models with operational geospatial processing in a fully open-source AI stack, ensuring transparency, extensibility, and digital sovereignty. Unlike proprietary solutions, the system enables public-sector–compliant deployment and long-term sustainability.

Added Value

The ac-URBAN solution delivers significant added value over existing EO analytics platforms by combining natural-language access, advanced geoprocessing, and a fully open-source architecture. In contrast to many competitors that rely on proprietary LLMs and closed geoprocessing engines, ac-URBAN is built on open technologies, enabling transparency, auditability, and long-term sustainability.

A key differentiator is the deep integration of AI agents with the open-source actinia and GRASS GIS ecosystem. Rather than producing black-box results, ac-URBAN translates user queries into explicit, inspectable process chains, allowing users and institutions to understand, validate, and reproduce analytical outcomes. This is particularly critical for public-sector, research, and policy-driven urban applications where explainability and trust are essential.

Unlike competing platforms that offer fixed products or opaque AI-driven outputs, ac-URBAN supports dynamic, user-driven analyses across global EO datasets while also enabling the integration of high-resolution, locally provided data. By leveraging open-source AI components, Retrieval-Augmented Generation, and cloud-native geoprocessing, ac-URBAN empowers non-technical users without sacrificing scientific rigor. The result is a vendor-independent, extensible solution that delivers transparent, adaptable, and locally actionable urban insights beyond the capabilities of proprietary, black-box alternatives.

Current Status

The implementation of the de-risking phase of the activity is from January to June 2026.