MMDA and the State Statistics Service of Ukraine explored using satellite data and geospatial products for official statistics. During the war, they provide an additional source of timely information.
The team presented a digital twin architecture and an interface for working with forest monitoring products. The development simplifies the use of satellite monitoring results.
In Corfu, the department presented research on tree cover mapping, field damage detection and forest biomass modelling. Prof. Shelestov and Prof. Kussul organised and chaired the thematic session.
Monitoring Ukraine’s forests in wartime is a major challenge: about 30% of forested areas are inaccessible to ground surveys because of occupation or proximity to the front line. Copernicus services do not cover Ukraine, and no national monitoring products or machine learning models adapted to Ukrainian conditions exist.
Project Goal: Develop methods for intelligent fusion of textual and geospatial information to create a digital twin of Ukraine's forest cover, supported by a cognitive interface built on agent-based AI technology. The system will provide analytical decision support for monitoring forest structural characteristics, change dynamics, and resource loss assessment.
The system will support regular monitoring, forest management planning, and official reporting on environmental damage. It serves government bodies, regional institutions, researchers, NGOs, and international partners, and it integrates Ukraine into European environmental monitoring systems.
Multi-Agent Framework for Scenario-Based Geospatial Modeling of Forest Biomass Dynamics under Natural and Anthropogenic Disturbances
Sofiia Drozd, Nataliia Kussul, Andrii Shelestov
16th International Conference Dependable Systems, Services and Technologies (DESSERT 2026), Corfu, Greece, October 2-4, 2026
From Land Cover Change to Geospatial Context: Assessment of Tree Cover Loss Statistics in Ukrainian Protected Areas During Russian Aggression from 2016 to 2025
Leonid Shumilo, Sofiia Drozd, Nataliia Kussul, Andrii Shelestov
16th International Conference Dependable Systems, Services and Technologies (DESSERT 2026), Corfu, Greece, October 2-4, 2026
Hybrid-Deterministic Fallback for Dependable LLM Provider Composition
Cherniatevych Anton, Drozd Sofiia, Kussul Nataliia, Shelestov Andrii, Giuliani Gregory, Poussin Charlotte
16th International Conference Dependable Systems, Services and Technologies (DESSERT 2026), Corfu, Greece, October 2-4, 2026
Agricultural Fields Damage Detection in Ukraine Using Sentinel-2 Data and ConvLSTM Neural Network Architecture with Attentions
Leonid Shumilo, Sofiia Drozd, Nataliia Kussul, Sergii Skakun, Andrii Shelestov
16th International Conference Dependable Systems, Services and Technologies (DESSERT 2026), Corfu, Greece, October 2-4, 2026
Annual Tree Canopy Cover Mapping Using GEDI and Multi-Source Satellite Embeddings: A Case Study for Ukraine
Bohdan Yailymov, Hanna Yailymova, Oleh Chaskovskyy, Oleksandr Matushevskyi, Andrii Shelestov, Nataliia Kussul, Svitlana Nosok, Mykhailo Kolomytsev
16th International Conference Dependable Systems, Services and Technologies (DESSERT 2026), Corfu, Greece, October 2-4, 2026
High-Resolution Tree Cover Density Mapping of Ukraine Using GEDI Lidar and Sentinel Imagery
Bohdan Yailymov, Hanna Yailymova, Nataliia Kussul
2026 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2026), Washington, DC, USA, August 9-14, 2026
Agentic AI for autonomous field damage risk areas detection and satellite data collection
Sofiia Drozd, Nataliia Kussul, Andrii Shelestov, Sergii Skakun
2026 IEEE International Geoscience and Remote Sensing Symposium (IGARSS 2026), Washington, DC, USA, August 9-14, 2026
Multi-agent Explainable AI Framework for Forest Disturbance Analysis and Biomass Recovery Modeling
Sofiia Drozd, Nataliia Kussul, Andrii Shelestov, Sergii Skakun
In: Nalepa, J., et al. Explainable AI in Space. EASi 2026. Communications in Computer and Information Science, vol. 3107. Springer, Cham, 2027.
https://doi.org/10.1007/978-3-032-36806-5_5Математичнi моделi оцiнювання щiльностi бiомаси лiсiв України на основi геопросторових даних
О. В. Матушевський, А. Ю. Шелестов
Теоретичнi i прикладнi проблеми фiзики, математики та iнформатики: матерiали XXIV Всеукраїнської науково-практичної конференцiї студентiв, аспiрантiв та молодих вчених (13-16 травня 2026 р., м. Київ, Україна). с. 347-351.
https://ela.kpi.ua/items/201b8daa-03d5-4138-8279-4a5409376a35MMDA and the State Statistics Service of Ukraine explored using satellite data and geospatial products for official statistics. During the war, they provide an additional source of timely information.
The team presented a digital twin architecture and an interface for working with forest monitoring products. The development simplifies the use of satellite monitoring results.
In Corfu, the department presented research on tree cover mapping, field damage detection and forest biomass modelling. Prof. Shelestov and Prof. Kussul organised and chaired the thematic session.
In Bremen, a multi-agent explainable AI framework for forest disturbance analysis was presented. It helps forecast biomass recovery from satellite observations.
Four department papers were presented in Washington, including AI agents and a high-resolution tree cover density map of Ukraine. Another work targets field damage risk areas.