“Advances in Artificial Intelligence for Aerospace Engineering” workshop
Mai 30th 2023 10:00-17:30 (same location-room than ODAS)
Context
In 2019 an agreement for cooperation in Artificial Intelligence (AI) was signed between DLR and ONERA. It has been decided to jointly consider AI research challenges in a number of application fields of particular importance in the aeronautics and space domains such as: robotics and autonomy in aerospace systems, hybrid simulation, digital transformation of aerospace industry, processing of geospatial information for earth observation. In 2020 the workshop to launch cooperation was organized in the form of a virtual conference that highlighted the interest of joint action of the two research institutions in the field of AI. Based on presentations of cooperation proposals, approaches and ideas were discussed in order to form the framework of the cooperation project. The participants identified areas of common interest for using AI. In 2021, a second workshop was also held virtually. Contributions from doctoral students and post-docs were presented. Researchers were able to discuss the use of AI in space robotics, for Earth observation, for improving the efficiency of computational fluid dynamics and for industry. Furthermore, advances and projects concerning the necessary robustness of AI in the aerospace context and its interaction with humans were exposed. In 2022, AI-related presentations were encouraged for ODAS. This led to the receipt of 18 presentation proposals on AI. Of these proposals, 7 were directly linked to the collaboration in AI.
Objectives and topics of the workshop
The main objective of the workshop is to share current advances of AI research work at DLR and ONERA through presentations in order to strengthen our collaborative work. The participants are doctoral students, post-docs and permanent researchers from the two institutions involved either in the development of techniques or in their application to the various fields targeted within the framework of cooperation. The workshop encourages exchanges about the use of AI in different areas such as: Space robotics, Earth observation, Computational fluid dynamics, Aerospace industry, Flight control, Mission management, Health monitoring and Maintenance. Nevertheless other application topics and fundamental AI topics are also welcome.
Organizing Committee
Robert Axmann : Robert.Axmann@dlr.de
Markus Kintscher: Markus.Kintscher@dlr.de
Philippe Bidaud : Philippe.Bidaud@onera.fr
Stéphane Herbin : Stephane.Herbin@onera.fr
Jean-Loup Farges : Jean-Loup.Farges@onera.fr
Contact
For DLR scientists, Markus Kintscher: Markus.Kintscher@dlr.de
For ONERA scientists, Jean-Loup Farges: Jean-Loup.Farges@onera.fr
Program
10:00 – 10:15 Informal welcome by Jean-Loup Farges (ONERA)
10:15 - 11:15 Session 1 chaired by Stéphane Herbin (ONERA)
10:15 - 10:30 Accelerating Materials Development and Testing with Autonomous AI-Powered Systems presented by Eric Breitbarth (DLR)
10:30 - 10:45 Artificial Neural Networks for Individual Tracking and Characterization of Wake Vortices in LiDAR Measurements presented by Niklas Louis Wartha (DLR)
10:45 - 11:00 Trends and perspectives on adversarial patch attacks presented by Pol Labarbarie (ONERA)
11:00 - 11:15 A VAE-based Method for Anomaly Detection in Multivariate Time Series presented by Baptiste Lambert (DLR)
11:15 - 11:45 Break
11:45 - 13:00 Session 2 chaired by Markus Kintscher (DLR)
11:45 - 12:00 FPGA based in-memory AI computing presented by Domenik Helms (DLR)
12:00 - 12:15 Semantic Mapping for Safe Autonomous Navigation of Ground Robots in Hazardous Environments presented by Quentin Serdel (ONERA)
12:15 - 12:30 Unsupervised Anomaly Detection for Space Gardening presented by Ferdinand Rewicki (DLR)
12:30 - 12:45 Optics/neural-network co-design presented by Marius Dufraissse (ONERA)
12:45 - 13:00 Data-based leakage detection in the manufacturing of large-scale CFRP components presented by Christoph Brauer (DLR)
13:00 - 14:00 Lunch break
14:00 - 15:30 Session 3 chaired by Philippe Bidaud (ONERA)
14:00 - 14:15 Autonomous and Reliable Aeroelastic System Identification using Density-Based Clustering, Gaussian Processes and Kalman Filtering presented by Robin Volkmar (DLR)
14:15 - 14:30 Crack detection on metallic materials using laser flying spot thermography and deep learning : a curriculum learning approach presented by Kevin Helvig (ONERA)
14:30 - 14:45 Multi term Adam optimizer for Physics Informed Neural Networks presented by Sai Karthikeya Vemuri (DLR)
14:45 - 15:00 Neural networks as a surrogate model for linear stability analysis of three-dimensional, compressible boundary layers presented by Paul Hoffmann (DLR)
15:00 - 15:15 Domain Incremental Learning for segmentation of remote sensing images. Preliminary analysis of the image encoder presented by Marie-Ange Boum (ONERA)
15:15 - 15:30 Development of Trustworthy Artificial Intelligence for AI-supported Air Traffic Controller Operations presented by Fotini Deligiannaki (DLR)
15:30 - 16:00 Break
16:00 - 17:15 Session 4 chaired by Robert Axmann (DLR)
16:00 - 16:15 Adaptive numerical dissipation control for large-eddy simulations of turbulent flows using artificial neural network models and a high order discretization scheme presented by Florian Setzwein (DLR)
16:15 - 16:30 Leveraging generative models to characterize image classifiers presented by Adrien Le Coz (ONERA)
16:30 - 16:45 Solver Consistent Data-Driven Correction of Coarse Grid CFD Simulations presented by Anna Kiener (DLR)
16:45 - 17:00 The Aircraft and its Manufacturing System: From Early Requirements to Global Design presented by Stéphanie Roussel (ONERA)
17:00 - 17:15 Data-driven turbulence modelling with the machine learning toolbox of the CFD solver trace presented by Alexander Bleh (DLR)
15 minutes slots for presentations should include some minutes for questions.
Practical informations
Despite originally planned on the afternoon only, the workshop will finally hold from 10h to 17h30.
Participants with no presentation should register on ODAS website(we need to track the number of participants).
For lunch, some restaurants are located in the neighborhood of Jusieu:
- Expansive
- KOKORO - 36 Rue des Boulangers
- Le Buisson Ardent - 25 Rue Jussieu
- Average
- L'Inévitable - 22 Rue Linné
- Les Sciences - 21 Rue Jussieu
- Cheap
- Mobster Diner - 20 Rue des Boulangers
- Heng Long - 45 Rue Linné
- Kumari - 22 Rue des Boulangers
- Chauds les Sandwichs - 4 Rue des Boulangers
- CALA Jussieu - 29 Rue Linné
- CASA DI PANINI - 17 Rue Jussieu
- Five Pizza - 41-43 Rue Linné
- Aux délices de Jussieu - 23 Rue Jussieu (only takeaway)
Remember that lunch break is only 13h to 14h
