Data on soil exploitation, on vegetation cover and on the transition between the different categories of use appear to be among the most asked information to formulate environmental and landscape heritage’s sustainable management strategies. They are used to check and verify the effectiveness of environmental policies and the integration of environmental instances in the sectorial policies (agriculture, industry, tourism, etc). Moreover, they are useful to define the framework in which to set an environmental monitoring: in the latter case more information can be required to verify specific impacts linked to the agronomic measures (fertilizations, plant-health measures, etc), the cultures protective effects in the different phenological phases, the different conditions connected to ground workings, cultural changing, meteorological events and irrigation.
Participants are asked to recognize agriculture (e.g. corn, radicchio and soy) for the years between 2015 and 2019 of field scale compatible with L5 Corine Land Cover level for an area subjected to pollution (ex: PFAS red zone), with the number of seasonal irrigation by using:
- the Optical Copernicus Sentinel-2 data and/or Radar Sentinel 1 Data;
- a given pilot area.
Some information for the calibration (some cultivated fields, and L5 Corine land cover structure) and some hour-scale daily rainfall from a reference station will be given.
Knowing the subtle material transport (sand, mud, silt, etc) in river realities, is important for hydrographic ponds management. Erosive and sedimentary processes of watercourse are important for hydro morphological aspects, fluviale ecology and for obtaining indication about changes of the river morphology over the years. Nowadays, also thanks to Copernicus program, many spatial high resolution satellite Images are available in order to be processed with specific algorithms to calculate parameters like total suspended matter (TSM), chlorophyll (CHL-a). Knowing how to recognize and distinguish hydric bodies in different hydrologic conditions is important in this framework.
Participants are asked to develop an arithmetic expression to define valid pixels to be used in SNAP software on the Sentinel-2 optical images in the C2RCC Processor to be able to recognize the main fluviale lattice in Veneto in a specific environment (active riverbed > 60m e.g: Po, Adige, Piave) both in murky and clean water.
Storm Vaia on 27-30 October 2018 happened because of a strong perturbation which brought persistent and extremely copious precipitations with strong winds and extreme gusts. The storm caused several hydrogeological instabilities, the fall of many trees, and the destruction of tens of thousands of alpine forests hectares.
Participants are asked to recognize the damaged areas and the largest number of landslides caused by Storm Vaia in a defined area in the province of Belluno, by using Unsupervised classification and SNAP software (by using Copernicus images before and after the storm).
Copernicus is the European Union Earth Observation Programme. A large amount of data are collected and used to give to providers, public authorities and international organizations a way to improve the life quality of citizens. The informations are free and open access. The collected data related to atmosphere monitoring, marine environment, land monitoring, climate change emergency and security.
The proposed challenge aims to rethink the standard building technology design approach in order to adapt in real-time to the forecasting given by Copernicus Data.
The objective is to find smart architectural solutions that improve the living performance of the users by responding to air quality, solar radiation, thermal variations and other climate change effects.
How would you dynamically interact with the built environment you are living in? How could Copernicus Data help us designing buildings with higher dynamic performances?
When satellites send data to Earth, they can be modified in order to create a wrong interpretation. In principle, who want to use these data for make prediction cannot be guaranteed of their authenticity.
Participants are asked to develop new algorithms – or optimize existing ones (like Patricia Trie or Merkle Tree )- in order to assure that data sent to Earth cannot be modified, so every user will be guaranteed of the originality of data and predictions made starting from these data doesn’t use incorrect data.
It is also possible to create a system which can identify suspicious patterns or corrupted data compared to historical patterns and data, by exploiting data collected from space, correlated with Big Data Analytics, Machine Learning and possibly other sources.
Let us suppose a user wants to use a GNSS (Global Navigation Satellite System), which provides a geospatial localization service. He expects that position, velocity and time provided by this system are correct and authentic.
Unfortunately there exist some techniques which an attacker can exploit in order to undermine data.
For example, cryptanalysis permits to defeat cryptographic defenses and thus falsify navigation data.
By using disclosure, an attacker can discover critical and sensitive information.
Spoofing permits to synthetize false signals with valid data in order to provide incorrect geospatial localization.
Last but not least, meaconing is the re-broadcast of the valid signals with delays.
Participants are asked to develop new algorithms, ore optimize existing ones (like TESLA protocol) in order to guarantee authenticity and correctness of localization provided by a GNSS.
Each of these threats can be performed by an attacker.
US government considers the energy sector as one of the sixteen vital infrastructure areas. In fact, we know what an extended blackout can cause to the entire system. On December 2015, Cyber hackers working for Russia paralyzed three Ukrainian services.
In energy sector, cyber threats can produce huge spending, exertions, interruption of services and financial/emotional influences on the business.
Participants are asked to develop innovative technologies, tools, and techniques (or optimize existing ones) to reduce risks of threats that can affect Nation’s critical energy infrastructure. An idea can be the development of an authentication algorithm to permit only to allowed people (for example who has credentials) the use of plants/machinery and data extraction.
Public safety can be affected by cyber threats.
Let us imagine what happens if the emergency number is hit by a denial of service attack.
Let us imagine what happens if there is a disclosure of information that Law Enforcement have to know.
Let us imagine what happens if a Fire departments dispatch system is intercepted.
Let us imagine what happens if a hacker takes control of an ambulance tracking system.
Participants are asked to develop new cybersecurity strategies (or improving existing ones) to protect communications, data and assets while maintaining high reliability and availability performance as well as interoperability between teams and organisations. An example of user categories include law enforcement, fire and emergency medical services actors involved in Public Safety.
Understanding customer behaviour is crucial in retail. However, datasets on customer journeys and shopping experiences are often so big, that companies struggle to analyse the information sufficiently. This can lead to a conversion rate of just 2-3% overall. E-commerce platforms could implement self-learning algorithms to correlate external factors, like satellite data about weather and the environment, with customer information like click data and search habits. Promotions could subsequently be tailored to each customer in real time. The customer experience would become more intimate, targeted and effective, and could lead to improved sales and brand loyalty. For example, AI could analyse georeferenced satellite data about detected cars in parking lots and merge it with past transactions, website searches and weather forecasts to predict future sales.
Use AI to analyse georeferenced satellite data, merge it with transactions, website searches and weather forecasts to predict future sales.
The electric utilities sector is involved with the generation and distribution of electricity. Machine learning is already used in several technologies that utilities use to manage the grid. Self-healing grids are able to move power around damaged equipment to keep customer lights on. In homes, consumer devices are able to react to human preferences and energy price signals to maintain comfort and control cost. The information collected through these technologies is essential to utilising AI and will lead to predictive grid maintenance, decreased power losses, and improved energy predictions.
The future electricity grid ought to be clean, seamless, cost-effective from generation to end- use, and capable of meeting capacity requirements. This would include:
• better management of clean energy sources, including renewables, natural gas and nuclear
• universal access to consumer participation and choice, including distributed generation, demand-side management and electrification of transportation
• holistically designed solutions, including regional diversity, hybrid AC-DC transmission and distribution systems, micro-grids, energy storage, and centralised-decentralised control
• Two-way flows of energy and information
Your challenge is to use AI and satellite technologies to add significant value by enabling accurate predictions of both energy demand and production, comprehensive monitoring and increased grid intelligent adaptability. For instance, renewable energy production can be better predicted through ultra-accurate weather forecasting; likewise, energy demand can be predicted based on consumers’ behaviours and external factors (e.g. weather).
AI would help to detect non-technical losses due to theft of power through correlating (suspicious) consumer’s behaviours patterns with contextual data (e.g.: presence of human activities). Finally, AI and Earth Observation weather data could enable preventive maintenance of power stations and transmission lines. This would reduce the time and costs associated with sending live-line workers out to check and fix remote grids.
Social good focuses on the utilization of AI to help achieve global sustainable development; it includes several application areas and is relevant to developed economies, economies in transition and for developing economies in particular. Some examples of AI for social good are: urban planning and development, environmental sustainability
Urban planning focuses on infrastructure, transportation, communication and distribution networks. AI technology can drive transformations in transportation infrastructure in urban areas, thereby improving mobility, safety and traffic flow. For instance, adaptive signal control systems are used quite commonly to control traffic lights. In the future, vehicle to Infrastructure (V2I) would allow vehicles to share information with components like RFID readers, cameras, traffic lights, lane markers, buildings, streetlights, signage and parking meters. V2I sensors integrated with AI could provide travellers with real-time advisories about such things as road conditions, traffic congestion, accidents, construction zones and parking availability. Likewise, traffic management supervision systems could use infrastructure and vehicle data to set variable speed limits and adjust traffic signal phase and timing (SPaT) to increase fuel economy and improve traffic flow.
Environmental sustainability is focused on education and conservation of endangered species and the sustainable management of ecosystems. Several efforts have been initiated to collect environment, weather and animal data provided by citizen volunteers, governmental and corporate initiatives. AI algorithms can be applied to merge in-situ data with relevant environmental data in order to gather valuable information to drive education and conservation activities at scale.
Your challenge is to combine AI and Copernicus dataset to develop an high value application in one of these fields: urban planning and development, environmental sustainability, circular economy