Category: Philab

Luxembourg-based company RSS-Hydro joins InCubed family

Posted on by gaiamaggi

With a project called FloodSENS, the Luxembourg-based company RSS-Hydro is now part of ESA’s InCubed programme. The project aims to develop an algorithm that efficiently identifies flooded areas under partial cloud coverage in optical satellite images, using Machine Learning and auxiliary high-resolution data from drones, digital elevation models, as well as water flow algorithms.

Floods are one of the most devastating natural disasters, accounting for the highest insured and uninsured losses annually, as well as costing many lives. The climate emergency intensifies the hydrological cycle. Consequently, the frequency and magnitude of extreme hydro-meteorological events, and therefore the risk of floods, are increasing, as confirmed to be happening in many places around the world. This flooding increase has devastating consequences, among which are a greater strain on humanitarian response efforts and the financial risk of the global (re)insurance market.

With the support of ESA InCubed programme, RSS-Hydro has now started to work on FloodSENS, a service that specifically addresses these issues by developing a flood mapping application for the open market that overcomes one of the major limitations of optical EO imagery during floods and is capable of efficiently scaling between optical satellite and drone images. Therefore this novel application would be able to efficiently reconstruct flooded areas under partial cloud cover in optical satellite images.

Guy Schumann, RSS-Hydro CEO, said: “This type of application is especially important for disaster response agencies at regional, national, and international level, who are keen to utilize the proliferation of open satellite data for flood mapping during emergencies. Additionally, in the insurance and re-insurance markets, stakeholders are interested in EO data to map the flood hazard of a high-impact event and, on a historical basis, to understand risk exposure and the changing nature of it.”

Bertrand Le Saux, ESA InCubed Technical Officer, added: “FloodSENS integrates both drone data and EO imagery with Machine Learning, to offer the best support to emergency response activities, as well as to the re-insurance market. At ESA, we see a big potential in this activity and we are ready to support it until its entry into the market”.

For Europe, but also for many other countries worldwide, future innovative EO-based and Machine Learning-powered apps would add considerable benefits to the existing products and services of the free Copernicus Emergency Management Service (EMS) and beyond. RSS-Hydro’s FloodSENS will position itself at the intersection of these two fields (EO technologies and Machine Learning application tools) and at the forefront of future EO-enabled innovative solutions, enabling a much more effective disaster response.

More information on the FloodSENS InCubed activity can be found here.

Image caption: The first half of 2019 was a devastating period for many countries in southeast Africa. After Cyclone Idai destroyed many places, in particular the port city of Beira, at the start of the year, Cyclone Kenneth ravaged northern Mozambique. Entire villages were destroyed and almost one million people were put at risk in the area. This partial cloud-free subset of a Sentinel-2 image from 3 May 2019 shows large areas under water in Pemba, regional capital of Cabo Delgado state, which experienced over 2 m of rain and flooding. FloodSENS will render optical imagery like this more usable during floods by reconstructing flooded areas under cloudy skies.

Links: FloodSENS, RSS-Hydro, Company Linkedin

OHB joins forces with ESA Φ-lab to boost the EO commercial market

Posted on by gaiamaggi

With the common objective to explore transformational technologies targeting the growing Earth observation commercial market, ESA Φ-lab and OHB have signed a Letter of Intent to define their basis for cooperation.

With the aim to define and develop transformative commercial services powered by Earth Observation (EO) space-based data and to foster the development of associated transformational technologies, OHB has reached an agreement with ESA Φ-lab to gain access to its unique capabilities and support. The cooperation may be extended in future to a joint research effort with OHB visiting researchers at the ESA Φ-lab.

“The commercial EO market has been following a fantastic evolution in the past decade, and it is essential for OHB to keep the pace and prepare solutions to the challenges of tomorrow. The cooperation with the ESA Φ-lab will foster the creation of transformative services representing a breakthrough to end-users.” says Sebastien Tailhades, Domain Manager Earth Observation at OHB.

Giuseppe Borghi, Head of the ESA Φ-lab, commented: “The ESA Φ-lab is the home of EO transformative technologies and commercial approaches within ESA. I strongly believe in compelling partnerships as a means to accelerate the EO future. We are very happy to add this new exciting collaboration to others already running at the ESA Φ-lab with industries, research centres and private investors. We look forward to transformative results!”

To know more: OHB, ESA Φ-lab

Primo Space Fund and ESA Φ-lab sign an agreement to boost New Space Economy

Posted on by gaiamaggi

Primo Space, a venture capital fund specialising in investments in the New Space Economy, and ESA have signed an agreement with the aim of joining efforts to support the growth of start-ups in the Earth observation Space Economy commercial sector. The collaboration will be coordinated by the Φ-lab, a division of ESA with a specific focus on fostering transformative innovation and its commercialization in the Earth observation segment.

The ESA Φ-lab mission is to accelerate the future of Earth observation. The lab also runs the commercial programme called Investing in Industrial Innovation, or InCubed, which is a 100+ M€ co-funding programme that focuses on developing innovative and commercially viable products and services exploiting the value of Earth observation imagery and dataset assets.

The activities of this collaboration between Primo Space and ESA Φ-lab range from exchanging information, organising events, highlighting common investment opportunities, to providing joint support to project activities partners and in general de-risking entrepreneur efforts and so increasing their success rate for both ESA InCubed and Primo Space Fund.

New Space is an extraordinarily dynamic sector of the space economy that is attracting the attention not only of industrial players but also of the investment community worldwide. Decision-makers worldwide are also increasing their attention to it. This is also happening because its effects can pervade almost every segment of economic activities and to some extent even everyday life.

“We believe that, in addition to making investments capable of creating value, which remains the primary objective of our activity, it is also important to contribute to the formation of a cohesive and determined ecosystem that favours the development of start-ups that create great innovations. This agreement goes exactly in that direction”, said the General Partners of Primo Space Fund – Matteo Cascinari, Raffaele Mauro and Giorgio Minola.

Giuseppe Borghi, Head of ESA Φ-lab, also commented: “The partnership with Primo Space Fund is a key step for us at ESA Φ-lab to support our commercial programme InCubed, which sees already 29 investments made in just 3.5 years (4 concluded and 3 already generating revenues) opening to Earth observation commercial entrepreneurs a privileged access to the private investors world to support their growth driven by disruptive innovation. Primo Space acts mainly in Italy, which is an InCubed key contributing country and we expect there a strong impact from this partnership.”

To know more: Primo Space Ventures, Primo Ventures LinkedIn

Get ready for the ESA-ECMWF workshop on Machine Learning for Earth System Observation and Prediction

Posted on by gaiamaggi

ESA and ECMWF are organising the second joint workshop on Machine Learning for Earth System Observation and Prediction that will be hosted from 15 to 18 November in an hybrid format. The call for abstracts is currently opened until 28 July. Participants registrations will start in August 2021. Submit your proposals.

The advantages of Machine Learning/Deep Learning (ML/DL) Artificial Intelligence techniques have been proven in a wide range of applications such as image recognition, traffic prediction, self-driving vehicles, and medical diagnosis. These techniques have also gained popularity within the Earth System Observation and Prediction (ESOP) community due to their ability to improve our understanding and prediction capabilities on the Earth’s complex and wide-scale dynamics.

Supported by the continuous increase in computing power, these techniques are valuable to automatically process and analyse a large range of available data but they still present some limitations, as for example the case of DL methods that need large amounts of curated and labelled data.

The workshop on “Machine Learning for Earth System Observation and Prediction” from 15 to 18 November, organised by ESA and ECMWF, will aim to demonstrate where and how this fusion between traditional ESOP techniques and new ML/DL methods reached a remarkable impact. It will also help to identify the remaining issues to be further explored. Presenters will demonstrate their contributions to this challenge and expand the discussion to provide a general overview of the subject.

After the first three days, covering the state-of-the-art, the working groups will discuss in parallel the current limitations and suggest how to advance so as to extract more value from this powerful fusion.

The output of the workshop is in the form of working group reports, which will be then summarised in a technical memorandum or a paper.

Abstracts can be submitted online until 28 July 2021. Participants registrations will start in August.

To know more: dedicated event website, ESA Φ-lab, ECMWF

ESA Φ-lab and IARAI announce their partnership

Posted on by gaiamaggi

The ESA Φ-lab and IARAI are proud to announce a new partnership supporting the growth of AI research to improve the understanding of our planet.

The Φ-lab division of Earth Observation Programmes Directorate at the European Space Agency (ESA) and IARAI are proud to announce a new partnership.

IARAI pursues basic and applied research in Artificial Intelligence (AI), from studying theory and white-boxing AI, i.e. gaining a quantitative understanding accessible by human intuition, to innovating for smart cities, sustainable mobility, climate change, and health. IARAI focuses on AI research for industrial scale real-world data to solve the major challenges of our society.

ESA’s mission is to develop European space capabilities for the benefit of all citizens. Φ-lab aims to accelerate the future of Earth observation by means of transformational innovations. Its AI4EO initiative promotes research in AI to exploit its potential with Earth Observation (EO) data.

IARAI and Φ-lab will now join efforts and resources to support the growth of AI research aimed to improve the understanding of our planet. The new partnership will bring joint research activities, organisation and promotion of scientific events, including conferences, competitions, workshops, and seminars, as well as focus group meetings and academic visiting programs, and other opportunities. We will collaborate to advance research in EO by applying and developing AI, and in particular Machine Learning (ML) methods for analysis and interpretation of big data on our planet.

As a first step in our partnership with Φ-lab, we are organising the 1st workshop on Complex Data Challenges in Earth Observation (CDCEO) 2021, in collaboration with the State Meteorological Agency of the Government of Spain AEMET and EODC. This workshop is held as a satellite event at the 30th ACM International Conference on Information and Knowledge Management (CIKM). The workshop focuses on application and development of advanced ML methods for effectively interpreting the high-dimensional heterogeneous data obtained by high-resolution EO missions. We invite all workshop participants to submit the extended versions of their accepted papers for publication a special issue of the IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. A special session of the workshop will present the winning solutions and highlights from a unique multi-sensor weather forecasting competition based on meteorological satellites data obtained in collaboration with AEMET and EUMETSAT’s Nowcasting Satellite Application Facility NWC SAF.

To know more: IARAI Press Release, IARAI website, Workshop on Complex Data Challenges in Earth Observation

ESA Φ-lab participates to ISPRS 2021 and IGARSS 2021

Posted on by gaiamaggi

The International Society for Photogrammetry and Remote Sensing Congress, or ISPRS, and the International Geoscience and Remote Sensing Symposium, or IGARSS, are two of the main annual rendez-vous events for the entire Earth observation community. This year the ESA Φ-lab has a significant  contribution to these events, taking part to sessions and scientific tracks and publishing several papers in both events.

The ISPRS Congress will be held virtually from 5 to 9 July.

The ESA Earth observation Programmes Directorate will be present with an online stand, a technology track titled “The ESA Earth Observation Platforms to leverage Open Science and Pioneer Innovative Applications” and two scientific tracks that will feature Φ-lab Research Fellows and Visiting Researchers:

  • Monday 5 July, 10:00 – 10:50, Jamila Mifdal and Nicolas Longépé will be amongst the speakers in th session titled “Thematic Information Extraction” (MO.3.2:#620)
  • Tuesday 6 July, 11:00 – 11:50, Dario Spiller will be amongst the speakers in the session titled “Hyperspectral Image Processing and Data Fusion” (TU.4.3: #539)

For more information: https://www.isprs2020-nice.com/programme/2021Digital_Program_06012021.pdf

The IGARSS event will be held from 12 to 16 July, also virtually.

The Directorate will be present with an online stand and some sessions, including one co-chaired by two ESA Φ-lab members, Nicolas Longépé and Bertrand Le Saux, titled “Physics-aware AI4EO and Simulation” on Friday 16 July, 14:40 – 16:10.

Ana Raquel Carmo and Nicolas Longépé will present their paper on “Deep Learning Approach For Tropical Cyclones Classification Based On C-Band Sentinel-1 Sar Images” on Friday 16 July, 8:30 – 8: 45, whereas other Φ-lab researchers, including James Wheeler, Dario Spiller and Alessandro Sebastianelli, also have successfully submitted accepted papers.

For more information: https://igarss2021.com/technical_program.php

Header image: Extreme weather events monitoring with AI4EO: tropical cyclone detection and categorization using deep neural networks. ©Raquel Carmo

To know more: ESA Φ-lab, ISPRS 2021, IGARSS 2021

InCubed supports InfoSequia’s satellite-based Drought Early Warning and Forecasting System

Posted on by gaiamaggi

ESA’s InCubed programme is supporting the development and pilot implementation of FutureWater’s new InfoSequia module, dubbed 4CAST. InfoSequia-4CAST combines historical and up-to-date observations of satellite-based meteorological and agricultural drought indices with climate variability indices to generate seasonal outlooks of water supply and crop yield failure alerts.

Water and food security are at risk in many places around the world, at present and even more so in the future, with significant economic and humanitarian consequences. Risk managers and decision-makers, such as water management authorities and humanitarian-aid agencies, can more effectively prevent harmful drought impacts if timely and actionable information is available on how water and food availability is affected.

FutureWater, a Dutch SME specialist in water management solutions, is developing InfoSequia-4CAST to provide users with timely predictions on drought impacts on crop yield and water supply. Seasonal outlooks are computed by a novel, state-of-the-art Machine Learning technique, previously  tested in applications for crop production forecasting and agricultural drought risk financing. The module responds to needs identified in previous applications and is a major extension of FutureWater’s Drought Early Warning and Forecasting System, InfoSequia. Another module, InfoSequia-MONITOR, is also being improved by incorporating satellite data to create multi-sensor drought indices.

Using multi-sensor, state-of-the art satellite data fully integrated with predictive models, InfoSequia-4CAST provides location-specific 3-6 month outlooks and warnings of crop yield and water supply failures to end users through a simple, intuitive user interface.

After signing  the contract with ESA in February 2021, FutureWater Project Manager Gijs Simons said: “This ESA supported InCubed activity is a unique opportunity for FutureWater to transfer a highly promising technology from the academic domain to a market-ready, operational decision-support tool. We are very enthusiastic about collaborating with ESA and a range of regional stakeholders to realise this product, which we believe can make a crucial difference for water management authorities and humanitarian NGOs worldwide.” 

FutureWater works with one of the world’s leading institutes in sustainability science, the Institute for Environmental Studies of the Free University of Amsterdam (IVM), to develop InfoSequia-4CAST. The product is targeted to the needs of water managers who intend to alleviate and mitigate the impacts of forthcoming drought periods by taking well-informed water management decisions, as well as humanitarian NGOs aiming to trigger ex-ante cash transfers with policyholders and farmer communities.

The InCubed activity includes collaborations with stakeholders and future end users in Spain, Colombia and Mozambique in order to establish user requirements, inform system design, and achieve pilot implementation of the system in the second project year.

Anna Burzykowska , ESA InCubed Technical Officer, added: ‘’The Earth Observation market sees an increased uptake of different type of products providing drought indices and early warning systems. The development of the FutureWater’s-4CAST module will enhance the current offering by producing advanced regional drought impact models with important improvements such as fusion with climate data records, yield forecasting as well as monthly and quarterly seasonal outlooks to more effectively manage the disaster impacts. We look forward to facilitating the project’s R&D activities as well as to supporting the engagement with new institutional and commercial customers.’’.

More information on the InfoSequia-4CAST InCubed activity can be found here.

To know more: FutureWater, InfoSequia-4CAST, InCubed activity, InfoSequia-MONITOR, IVM

The next ELLIS-ESA workshop on Quantum Algorithms and Machine learning is coming

Posted on by gaiamaggi

ELLIS and ESA Φ-lab are co-organising the next workshop on Quantum Algorithms and Machine Learning for huge data analysis, simulation and potential Earth observation applications on Thursday 27 May, from 13:00 to 19:00 CEST. Registrations are available through the workshop website.

Quantum Computing has the potential of revolutionising information processing and becoming a key enabler for computationally hard engineering and scientific problems. Recent advances in quantum technologies and quantum algorithms make this game-changing turn more likely.

Quantum computers promise to redefine computing and allow to make certain complex computations in a drastically reduced time. In terms of data analytics, they might enable us to sample and explore huge volumes of scattered data, to identify and retrieve specific patterns, and optimize functionals for many kinds of use-cases. In terms of physics simulation, they will enable solving inverse problems entailing partial differential equations with applications to e.g. geophysical fluid dynamics. Today, thanks to progress in Noisy Intermediate-Scale Quantum (NISQ) devices and classical-quantum hybrid computing, promising approaches are already tested on real machines.

Earth observation (EO) gathers global information about our planet’s physical, chemical and biological systems via sensing devices. A recurrent issue in EO is the solution of inverse, ill-conditioned problems, which includes specific land-cover identification, biophysical parameter estimation and feature extraction, atmospheric inverse problems, gravimetry, etc. EO needs the unprecedented power of Quantum Computers to face computing challenges such as those in:

  • Synthetic Aperture Radar (SAR) processing: phase unwrapping for SAR, polarimetric SAR and interferometric SAR, with applications to elevation modelling for SAR imaging;
  • Multispectral / hyperspectral processing: machine learning applied to images and hyperspectral data cubes, resulting in a new range of AI4EO methods, optimised and  scaleable to the challenging volumes of global EO daily imaging and archives, key to all downstream products;
  • Processing of Earth system measurements (aerosols, atmospheric or ocean measurements, etc.) in climate and weather modeling and data assimilation, with better solvers and optimisers, leading to improved temporal modelling, forecasting and Earth system simulation.

The programme will include:

  • 13:00-17:00 CEST – Keynote talks (with audience)
  • 17:00-19:00 CEST – Roundtable and panel discussion (closed session)

Useful links: ELLIS-ESA workshop website, Workshop registration form, ELLIS QPhyML Program

Φ-week 2021 – Save the Date

Posted on by adeluca

The European Space Agency is organising the fourth edition of Φ-week on 11–15 October 2021. This year main theme will be the Earth Observation New Space economy and its associated innovations. The event will be virtual and free-to-attend for the public and in person for the invited speakers. The calls for proposals for Side Events, e-Posters and the e-Exhibition are already opened.

The fourth edition of ESA Φ-week will be organised virtually from 11 to 15 October 2021 and in person for invited speakers, COVID permitting. As in the past editions, this event will focus on how to accelerate the future of Earth Observation (EO), on presenting recent developments in EO Open Science and latest trends in EO markets, on exploring bold and transformative ideas that ESA’s Φ-lab and Data Applications Division support and scale up along with researchers, start-up’s, industry and private investors.

The main theme of Φ-week 2021 is the New Space economy and associated innovations. The sessions, posters and side-events will highlight how the New Space economy is developing in Europe and alongside competition worldwide, and how it contributes to the EU Green Deal, Digital Europe Programme, Destination Earth initiative, UN SDGs, and in general to the EU Space Strategy and the European space sector.

EO New Space is indeed a global trend of emerging investment and entrepreneurial philosophy that, together with key technological advancements, is enabling a private space industry largely driven by commercial motivations and will eventually evolve into EO Commercial Space. This trend is transforming several space economy sectors, for example providing small and recoverable launchers making space more accessible, delivering rich and affordable information from space, and providing internet access worldwide based on satellite constellations.

Φ-week will include inspiring talks, key sessions, roundtables, side events and other initiatives that aim to connect a multi-disciplinary community, including EO downstream and upstream world market leaders, researchers, Earth scientists, non-space companies, technology leaders, entrepreneurs, start-up’s and innovators, New Space operators, private investors, ICT players, ESA, Member States and EC representatives.

The calls for proposals for Side Events, e-Posters and the e-Exhibition are opened until 31 May and available here.

The detailed programme and all relevant information and updates will be published in the coming days on the official website: phiweek.esa.int/

Stay tuned!

Useful links: Φ-week 2021 website, Relive Φ-week 2020

AI4EO launches its first Challenge to improve Air Quality & Health

Posted on by adeluca

AI4EO is an ESA initiative from the Φ-lab that aims to bring together the worlds of Artificial Intelligence (AI) and Earth Observation (EO) to foster interaction and collaboration. The initiative will include various Earth Observation challenges, created by the Φ-lab, to address important issues using AI and adopt the best solutions. Register for the first AI4EO Challenge on Air Quality and Health now.

AI4EO aims to act as a bridge between EO and AI. The initiative achieves this through various challenges, an ambitious and wide-ranging community, social media, and networking campaign. This ensures the long-term future of the initiative, and encourages interaction between EO and AI to solve important issues.

The AI4EO Project Team is currently organising several thematic Challenges where participants will be called upon to find solutions to important issues using AI on EO data.

The first AI4EO challenge, the Air Quality and Health Challenge, emerges from the need expressed by ECMWF and the Copernicus Atmosphere Monitoring Service (CAMS) for higher spatial resolution of air quality data and products, using EO data from Copernicus Sentinel-5P with AI technologies.

Better information about air pollution and reducing emissions of key pollutants such as fine particulate matter and nitrogen dioxide could save millions of lives. The objective of this challenge is to downscale air quality products such as Particulate Matter (PM2.5) and Nitrogen Dioxide (NO2) to a resolution that can be used on a local level. Participants will produce AI-powered downscaling methods on three areas of interest: North Italy, California and South Africa. The most successful methods could be considered for benchmarking current methodologies. 

Participants will work in teams to address the AI4EO challenges and four winning teams, including a student team, will be awarded ‘’AI4EO points’’ to spend on various prizes provided by sponsors and implementation partners. The challenge was launched at the beginning of February 2021 and will remain open until 15 May 2021. Anyone with an interest in AI and EO can register for the challenge and join a team. More information on the Challenge can be found here.

More information on AI4EO can be found here.

To know more: AI4EO, Air Quality and Health Challenge, ECMWF, CAMS