Harnessing AI for Climate Resilience: The Role of UNIVPM and AMAP in the AINATURE Project

Climate change, land degradation, and water pollution represent increasingly urgent challenges for the Adriatic-Ionian region. In response, the AINATURE project “Exploitation of AI for the protection of natural resources in the ADRION region”, has united research institutions, regional authorities, and innovation agencies from across the ADRION area to build an integrated, AI-powered platform for the early detection and prevention of environmental crises.

At the heart of AINATURE are two Italian partners whose complementary expertise plays a crucial role: the Marche Polytechnic University – UNIVPM (Department of Industrial Engineering and Mathematical Sciences – DIISM Department) (Fig. 1a) and the Agency for Innovation in the Agri-food and Fisheries sectors (AMAP) (Fig. 1b). Together, these institutions will collaborate with the aim to apply cutting-edge artificial intelligence and satellite-based technologies to monitor and mitigate soil erosion and water pollution.

Figure 1: the two Italian partners: the Marche Polytechnic University – UNIVPM (Department of Industrial Engineering and Mathematical Sciences – DIISM Department)  and the Agency for Innovation in the Agri-food and Fisheries sectors (AMAP).

A Transnational Vision for Environmental Protection

AINATURE addresses several pressing environmental threats, in particular the increasing frequency of floods and wildfires, often exacerbated by climate change, and water pollution (examples in Fig. 2). The project capitalizes on the knowledge developed in several Horizon Europe and H2020 projects (including TREEADS, Silvanus, PathoCERT, aqua3S), and builds upon AI tools capable of analyzing freely available satellite imagery and crowdsourced data via a de

dicated mobile application.

Figure 2: extreme drought, hot temperature, and gusting winds have set the stage for destructive wildfires in the state of New Mexico in June 2025 (a); the striking contrast between the muddy head of the King Sound – gulf of the Indian Ocean, northwestern coast of Western Australia – and the blue waters as a result of sediment transport (b) (image credits: MODIS Land Rapid Response Team, NASA GSFC).

This approach introduced by the project aims to contain costs compared to conventional monitoring systems based on physical sensors, drones, and field measurements. By integrating AI-driven data that are being processed in real time with Big Data analytics, AINATURE aims to offer a flexible, scalable, and sustainable monitoring system for the entire ADRION region.

UNIVPM: Advancing AI to Monitor Soil Erosion

Within the Italian context, UNIVPM through the DIISM leads on the technical and methodological development for applying AI models to soil erosion assessment to support rural lands. Leveraging its strong research tradition in AI, engineering, and environmental monitoring, UNIVPM focuses on developing monitoring tools to model soil erosion processes and water pollution and foresee erosion risks that may directly impact water quality downstream.

Soil erosion processes are being investigated within the selected pilot area by AMAP. The pilot area is located within the Esinante river watershed near Apiro and Sant’Urbano within the province of Ancona, in the Marche region (Fig. 3). This agricultural, hilly and clay-rich terrain can be subject to significant soil and water losses particularly during heavy rainfall events. Such processes not only diminish soil fertility but also transport suspended sediments and nutrients (notably phosphorus and nitrogen) into nearby rivers..

Figure 3: location of the pilot area (yellow waypoint) within the context of the Esinante river watershed.

UNIVPM’s research team, in collaboration with AMAP and their stakeholders, aim to implement an innovative, field-based monitoring system through capitalization of projects results of the previous years. Custom-designed equipment collectors will collect run off water and solid sediments after each rainfall event, while detailed pedological and chemical analyses will provide further information on erosion dynamics, organic carbon loss, and nutrient depletion.

AMAP: Bridging Research and Practice in Sustainable Agriculture

As the operational agency of the Marche Region for agri-food and fisheries innovation, AMAP brings to AINATURE its extensive experience in climate adaptation, soil protection, and nitrate management. In particular, AMAP will deliver the data collected of the pilot area to the AI tools developed by UNIVPM with its ongoing agro-climatic and pedological monitoring activities.

Through its network of ago-meteorological stations, soil mapping services, and participatory stakeholder engagement, AMAP will contribute to provide with essential data inputs on soil quality, land use practices, and agricultural nutrient cycles. This informa

tion will enable the AI system to refine its predictive capabilities and to generate actionable recommendations for farmers, policymakers, and regional authorities.

In the Marche pilot case, AMAP focuses on monitoring the impact of agricultural practices on soil erosion, water, nitrate and phosphate losses. This synergy between AI-driven modelling and practical agri-environmental management underpins the region’s efforts to align with EU Green Deal objectives and the Do No Significant Harm (DNSH) principle.

The Power of Synergy: UNIVPM and AMAP Collaboration

The cooperation between UNIVPM and AMAP showcases an exemplary model created within EU project, between university-agency partnership, in which scientific excellence merges with applied innovation and regional policy. While UNIVPM develops the core AI models, AMAP ensures data continuity, stakeholder involvement, and policy relevance.

Together, they ensure that the outcomes of AINATURE are not only scientifically robust but also operationally viable for replication across other ADRION regions.

A Future-Oriented Ai-Nature Digital Platform

The AI models to be developed for the Marche pilot area will feed into the AINATURE Platform, an integrated digital environment developed under the technical coordination of CERTH (Greece). This platform will host AI-powered early warning systems for floods, wildfires, and water pollution, accessible to all partners and stakeholders in the ADRION region.

Ultimately, AINATURE is more than a collaborative project; it is a forward-looking initiative that demonstrates how artificial intelligence, combined with transnational cooperation and regional know-how, can build climate resilience, protect natural resources, and foster sustainable development across borders.