• Hackathons
  • Features
  • Blog


Machine Learning and Remote Sensing to improving crop productivity and water use efficiency

  • 101 Raised
  • 1 Juries


  • Italy


  • 3. Protecting our critical infrastructure



Water Wise is a DSS (Decision Support System) Machine Learning-based solution that aims to improve crop productivity and water use efficiency thanks to remote sensing and machine learning. 

This is because water is one of the critical infrastructures according to the Ministry of the Interior. Water shortages could erupt into socio-political tensions and endanger security and biodiversity.

In particular Water Wise applies advanced machine learning algorithms, such as deep neural networks, to extract information from the satellite data and to model the non-linear and heterogeneous relationships between the indicators and the in-situ measurements of crop yield. 

Water Wise helps to prevent farmers monitor and managing crop water stress and irrigation scheduling in real-time, saving water, energy and costs while increasing yield and quality. WaterWise is a game changer in the field of precision agriculture and sustainable water management.

CropSense leverages EU space technologies such as Sentinel-1 and Sentinel-2 satellites to collect high-resolution data on various indicators of crop water status and performance, such as leaf area index, soil moisture, canopy temperature, NDVI, evapotranspiration and crop water productivity.


Water is one of the critical infrastructures according to the Italian Ministry of the Interior. Water shortages could erupt into socio-political tensions and endanger security and biodiversity. Water is also a public good that must be protected by individual citizens and at an international level. Therefore, it is essential to optimize the water management process for saving water resources and ensuring their fair and sustainable use. Using space technology aims to achieve this goal. Our solution can help decision-makers and stakeholders in the water sector to plan and implement effective water policies and practices that can enhance water security and resilience.


The product aims to use a mix of data: in particular, it will use Sentinel 1, Sentinel 2, Sentinel 3, weather data, and data from soil humidity sensors to train our Deep Neural Network (DNN) Model.

Once trained, the neural network will use real-time satellite data to output how much water a soil needs, in order to optimize water management.

We want to exploit the advantages of each Sentinel: thanks to Sentinel 1, service continuity is guaranteed, being an imaging radar mission that provides continuous images for all seasons, day and night in the C band: with this, we measure the humidity of the soil.

Sentinel 2 provides high-resolution optical images and some vegetation indices such as NDVI or LAI are measured.

By integrating the Sentinel 3 data, the value of evapotranspiration can also be obtained, i.e. the sum of water vaporised into the atmosphere by direct evaporation from the soil and by transpiration from plants: when there is drought, this value is higher than the water available in the soil.

The data from the soil moisture sensor are useful as ground reality and the meteorological data are used to monitor rainfall and atmospheric effects, useful for water management and saving water resources.


We address two different market segments with our solution. 

The first segment is local and regional authorities, the government and the ministry for agriculture. They will benefit from using our system to avoid social and political tensions caused by water scarcity and mitigate them using active initiatives for water management and preservation. Moreover, adopting water management strategies would help to release pressure on the hydric infrastructure. This segment represents a large and stable market with a high potential for growth and social impact. Our value proposition for this segment is to provide a comprehensive and reliable DSS on large scale and resolutions that can help them optimize water allocation and distribution, forecast water demand and supply, and evaluate the environmental and economic impacts of different scenarios.

The second segment is big farmer companies and consortia, which would benefit from using our system to improve crop productivity. This segment represents a competitive and dynamic market with high demand for innovation and efficiency. Our value proposition for this segment is to provide a customized and flexible DSS with high resolution and on a small scale that can help them improve irrigation scheduling and prediction for crop productivity. Our system can help them reduce water and energy costs, comply with environmental regulations, and enhance their reputation as sustainable and responsible businesses.

Our competitive advantage in both segments is our use of advanced machine learning techniques that can handle complex and uncertain data, and learn from different sources, in particular exploiting the capabilities of EU infrastructure for satellite data.


Four people with different backgrounds:

- Claudio Caporusso, Data Scientist

- Angela Dimita, Marketing Specialist

- Anna Verlanti, Earth Observation Specialist

- Akhilesh Kola, Machine Learning Expert