Powered by Indico
v3.2.9
indico will be upgraded to the latest version on Tuesday 30th July. It may be unavailable all day.
The Software Systems Division (TEC-SW) and Data Systems, Microelectronics & Component Technology Division (TEC-ED) Final Presentation Days are scheduled to take place on Wednesday 30 November & Thursday 01 December 2022 via WebEx/ESTEC (TBC)
All material presented at the Final Presentation Days must, before submission, be cleared of any restrictions preventing it from being published on this web-site.
More info will become available beginning of November 2022
With the advantages and appealing performances of artificial Intelligence (AI) in different applications, space scientists and engineers have shown great interest in AI-based solutions to space scenarios. However, different from terrain applications, the decision of the space vehicles offered are critical and should be trustable in the uncontrolled and risky environment, resulting in the most significant challenge in the use of AI-based techniques for space missions of acting sensibly for the unanticipated and complex situations. We, therefore, study the Explainable Artificial Intelligence (XAI) techniques that are potentially applicable to software-based GNC scenarios, including relative navigation for spacecraft rendezvous, crater detection and landing on asteroids or the Moon. The explainable tools related to the XAI algorithms are developed to make onboard intelligent techniques transparent to ensure a trustable decision while meeting the level of performances required by the space applications within an uncertain and acceptable boundary level. A comprehensive framework is proposed to address the XAI-base space software for spacecraft GNC systems, including syntenic dataset generation, relative navigation scenario building, XAI model developing, software verification and testing, etc.
Artificial Intelligence methods such as Deep Neural Networks can provide performances that surpass those of more classical techniques. The potential improvements do not come for free, as the difficulties with these techniques lies in the explainability of their results. For this reason, introducing them into critical embedded systems poses some challenges. In this work, result of an ESA project, we analyse two use cases where Deep Neural Networks are introduced into vision-based navigation systems and implement them on representative space avionics. We developed one model for each use case, the first aiming at detecting and localizing craters in a descent and landing phase over the Moon, and the second aiming at identifying previously selected patches of an unknown satellite on a life visual camera feed.
The use of Artificial Intelligence (AI) has been recognized as a major advance in several industries including Automotive, Agriculture and Healthcare, disrupting traditional approaches and leading to a myriad of novel applications. The space domain has also been reached by the innovation potential of AI, chiefly in Earth Observation applications. Moreover, the increasingly more powerful processing units, together with enhanced and less computationally intensive AI algorithms, make it possible to explore new AI applications, especially for onboard implementations. The objective of the AIVIONIC technology development project is to implement a HW/SW demonstrator of an AI-based Visual Navigation System that can be applied to mission scenarios of Lunar landing and rendezvous and capture of a non-cooperative target. This follows a novel development line towards demonstrating the use of AI in space critical systems, in a dependable manner.
Satellites have become an essential part in our lives: from navigation and weather forecasting to broadcasting and disaster management. They acquire enormous amounts of data which are sent back to Earth for further processing and analysis. Cost-effective downlink with high throughput is usually not available, especially for small satellites. Although satellites are collecting large amounts of data from various sensors, the downlink capacity is not sufficient enough to bring data to Earth in a timely and cost-effective manner. Another problem is limited resources on satellites (power, space, memory, etc.) and their usage has to be carefully designed and optimized. However, space environment can be very unpredictable and not easily simulated on Earth. Constant auto-optimization by “learning” from the measured data could help to utilize the resources in the optimal way.
The above mentioned problems can be solved with an AI module with onboard processing capability. For example, if a satellite is monitoring objects of interest on the ground, defocused or cloud-filled images are useless and do not need to be downstreamed. Moreover, objects of interest could be detected and cropped or parameterized from images onboard (e.g. regions of interest like agricultural land, urban regions, etc.), further reducing the amount of data for downstreaming. Another example is battery usage optimization where an onboard module could monitor battery power and learn how to reduce energy consumption or manage power distribution, prolonging the designed lifetime. Many of these problems could be solved by AI algorithms that would filter out unwanted images and would reduce the amount of data that needs to be downlinked or by optimizing satellite behavior like energy consumption, navigation, etc.
An ITAR-free fully integrated radiation-hard time-to-digital converter (TDC) with a sub-10 picosecond single-shot precision is developed by Magics Technologies in the scope of an ESA contract. The design is successfully integrated on an ASIC and supports high resolution, high accuracy and high precision at low power with zero dead-zone in a dynamic measurement range from 0s up to more than 3s. The ASIC is validated for the full temperature of -40°C to 125°C.
CCSDS EDS Pink Sheet Interoperability and Tooling (ESA Contract No. 4000141831/23/NL/AS)