ESA’s 2020 Φ-week occasion kicked off this morning with a sequence of stimulating speeches on Digital Twin Earth, updates on Φ-sat-1, which was efficiently launched into orbit earlier this month, and an thrilling new initiative involving quantum computing.
Digital Twin Earth
The third version of the Φ-week occasion, which is completely digital, focuses on how Earth remark can contribute to the idea of Digital Twin Earth – a dynamic, digital duplicate of our planet which precisely mimics Earth’s conduct. Continuously fed with Earth remark knowledge, mixed with in situ measurements and synthetic intelligence, the Digital Twin Earth offers an correct illustration of the previous, current, and future adjustments of our world.
Digital Twin Earth will assist visualize, monitor, and forecast pure and human exercise on the planet. The mannequin will be capable of monitor the well being of the planet, carry out simulations of Earth’s interconnected system with human conduct, and assist the sphere of sustainable improvement, due to this fact, reinforcing Europe’s efforts for a greater surroundings in an effort to reply to the pressing challenges and targets addressed by the Inexperienced Deal.
At the moment’s session opened with inspiring statements from ESA’s Director Common, Jan Wörner, ESA’s Director of Earth Remark Programmes, Josef Aschbacher, ECMWF’S Director Common, Florence Rabier, European Fee’s Deputy Director Common for Defence Trade and House, Pierre Delsaux, in addition to Director Common of DG CONNECT on the European Fee, Roberto Viola.
Pierre Delsaux commented, “As our EU Fee President repeated just lately throughout her State of the Union speech, it’s clear we have to tackle local weather change. The Copernicus program provides us a number of the greatest devices, satellites, to present us a whole image of our planet’s well being. However house isn’t solely a monitoring software, it is usually about utilized options for our financial system to make it extra inexperienced and extra digital.”
Roberto Viola stated, “Φ-week is the week for disruptive expertise and it’s communities like this that our European programmes had been designed to assist.”
Florence Rabier added, “Machine studying and synthetic intelligence might enhance the realism and effectivity of the Digital Twin Earth – particularly for excessive climate occasions and numerical forecast fashions.”
Jan Wörner concluded, “Φ-week is the right instance of the New House method specializing in disruptive innovation, synthetic intelligence, agility and adaptability.”
Throughout the week, consultants will come collectively to debate the position of synthetic intelligence for the Digital Twin Earth idea, its sensible implementation, the infrastructure necessities wanted to construct the Digital Twin Earth, and current concepts on how industries and the science neighborhood can contribute.
Earlier this month, on three September, the primary synthetic intelligence (AI) expertise carried onboard a European Earth remark mission, ɸ-sat-1, was launched from Europe’s spaceport in French Guiana. An enhancement of the Federated Satellite tv for pc Methods mission (FSSCat), the pioneering synthetic intelligence expertise is the primary experiment to enhance the effectivity of sending huge portions of knowledge again to Earth.
At the moment, ESA, together with cosine distant sensing, are glad to disclose the primary ever hardware-accelerated AI inference of Earth remark photographs on an in-orbit satellite tv for pc – carried out by a Deep Convolutional Neural Community, developed by the College of Pisa.
ɸ-sat-1 has efficiently enabled the pre-filtering of Earth remark knowledge in order that solely related a part of the picture with usable info are downlinked to the bottom, thereby enhancing bandwidth utilization and considerably decreasing aggregated downlink prices.
Preliminary knowledge downlinked from the satellite tv for pc has proven that the AI-powered automated cloud detection algorithm has accurately sorted hyperspectral Earth remark imagery from the satellite tv for pc’s sensor into cloudy and non-cloudy pixels.
Massimiliano Pastena, Ф-sat-1 Technical Officer at ESA, commented, “We’ve got simply entered the historical past of house.”
At the moment’s profitable utility of the Ubotica Synthetic Intelligence expertise, which is powered by the Intel Movidius Myriad 2 Imaginative and prescient Processing Unit, has demonstrated actual on-board knowledge processing autonomy.
Aubrey Dunne, Co-Founder and Vice President of Engineering at Ubotica Applied sciences, stated, “We’re very excited to be a key half of what’s to our information the primary ever demonstration of AI utilized to Earth Remark knowledge on a flying satellite tv for pc. This can be a watershed second each for onboard processing of satellite tv for pc knowledge, and for the way forward for AI inference in orbital purposes.”
As the general 2017 Copernicus Masters winner, FSSCat, was proposed by Spain’s Universitat Politècnica de Catalunya and developed by a consortium of European corporations and institutes together with Tyvak Worldwide.
Additionally talked about in his opening speech this morning, Josef Aschbacher made a particular announcement relating to an thrilling new ESA initiative, the “EOP AI-enhanced Quantum Initiative for EO – QC4EO” in collaboration with the European Group for Nuclear Analysis (CERN).
Quantum computing has the potential to enhance efficiency, lower computational prices and clear up beforehand intractable issues in Earth remark by exploiting quantum phenomena resembling superposition, entanglement, and tunneling.
The initiative entails making a quantum functionality which may have the flexibility to unravel demanding Earth remark issues by utilizing synthetic intelligence to assist programmes resembling Digital Twin Earth and Copernicus. The initiative will probably be developed on the Φ-lab – an ESA laboratory at ESA’s heart for Earth remark in Italy, which embraces transformational innovation in Earth remark.
ESA and CERN take pleasure in a long-standing collaboration, centered on technological issues and elementary physics. This collaboration will probably be prolonged to hyperlink to the CERN Quantum Know-how Initiative, which was introduced in June 2020 by the CERN Director Common, Fabiola Gianotti.
Via this partnership, ESA and CERN will create new synergies, constructing on their widespread expertise in massive knowledge, knowledge mining and sample recognition.
Giuseppe Borghi, Head of the Φ-lab, stated, “Quantum computing along with AI are maybe probably the most promising breakthrough to return alongside in pc expertise. Within the coming years, we are going to see extra Earth or house science disciplines using present or future quantum computing strategies to unravel geoscience issues.”
Josef Aschbacher added, “ESA will exploit the broad vary of specialised experience accessible at ESA and we are going to place ourselves in a singular place and take a number one position within the improvement of quantum applied sciences within the Earth remark area.”
Alberto Di Meglio, Coordinator of the CERN Quantum Know-how Initiative, stated, “Quantum applied sciences are a quickly rising discipline of analysis and their purposes have the potential to revolutionize the way in which we do science. Getting ready for that paradigm change, by constructing information and instruments, is important. This new collaboration on quantum applied sciences bears nice promise.”