These innovations in GeoAI open up new possibilities in the analysis of huge volumes of geospatial data and the processing of advanced spatial analyses, making it possible, for example, to refine weather forecasts, improve the quality and reliability of satellite imagery, and link advanced language models with geographic information systems (GIS). The innovations brought about by the collaboration of the PoliRuralPlus and ALIANCE projects strengthen the Czech Republic's leading position in Earth Intelligence and future technologies.
The PoliRuralPlus and ALIANCE projects open the door to new collaborations and strengthen links between European research organisations and the professional community in agriculture and spatial planning. By combining the latest technologies such as artificial intelligence, machine learning and remote sensing, they are delivering breakthrough results that enable more efficient land planning, better use of resources and data-driven decision-making.
To automate video editing, Justýna Frommová used a neural convolutional network (EfficientNet) with a classification head that can recognize the beginnings and ends of exchanges based on frame analysis. Based on this data, she created a program that defines time tracks and automatically cuts the video in production according to them.
They also point to new opportunities to invest in the development of innovative solutions in geospatial data, artificial intelligence and their applications in precision agriculture. This sector represents a modern way of farming that brings economic profit to farmers while responding to the new demands of society and the technical possibilities of the digital age. These technologies thus lead to greater efficiency, sustainability and strategic decision-making at all levels, while promoting economic growth and sustainable use of natural resources.
Key project deliverables:
-
More accurate weather forecasts: integrating AI-powered global models (models that predict weather around the world) with data from local weather stations has led to a significant increase in the accuracy of weather forecasts, which will positively impact agriculture, crisis management and other public service sectors.
- Removing clouds from satellite imagery: Advances in Sentinel 2 data processing have made it possible to remove clouds and improve the reliability of imagery, contributing to more effective environmental monitoring, tracking land-use change and managing responses to natural disasters.
-
Linking large language models to GIS: This link enables more detailed analysis of spatial data, crucial for infrastructure planning, improving town services and decision-making in public administration. The PoliRuralPlus project integrates geospatial predictors with LLMs such as ChatGPT, improving the analysis of local conditions and the prediction of geographic change. The Hub4Everybody platform provides a versatile tool for creating knowledge hubs supported by spatial data and digital communities.
Planned development and cooperation
The PoliRuralPlus project will continue to develop and scale up GeoAI applications in practice for effective planning, strategic decision-making and support for sustainable development. It will also benefit from linkages with other European projects under Horizon Europe and ERASMUS programmes, enabling further development of innovative applications with global reach. The main contribution of the ALIANCE project is the Data Integration Platform, which provides efficient data management and uses advanced AI technologies to reconstruct missing satellite imagery data, ensuring farmers have accurate information even in adverse conditions.
More information about the PoliRuralPlus project https://www.poliruralplus.eu/
More information about the ALIANCE project https://aliance.hsrs.cz/
