Sending a rocket into space is tricky, right? Well, yes and no. <\/p>\n\n\n\n
If you’re a rocket scientist, combining liquid oxygen and liquid hydrogen to generate thrust to propel a rocket may seem deceptively straightforward. However, the extreme conditions generated and where they operate dictate the need for precise and robust controls. This is challenging to overcome. Amongst the most critical tasks is ensuring that engine system faults are detected quickly and reliably.<\/p>\n\n\n\n
This sounds complex, and truthfully, it is! Thankfully, the power behind machine learning and AI is changing how this is done.<\/p>\n\n\n\n
The minds behind Proekspert’s Data Science team, comprised of Alfonso Garc\u00eda-Sosa, Karl Kaspar Haavel, Henry Ginter, and Laura P\u00e4eske, are implementing Transformer Autoencoder anomaly detection and classifications<\/strong> on dozens of complex sensors required, including:<\/p>\n\n\n\n The Engine Health Monitoring System developed by Proekspert makes it possible to follow and detect faults before they transpire and within fractions of seconds on the edge chip, a device on-board the space rocket. <\/p>\n\n\n\n Such a monitoring system makes diagnosing and reconfiguring engine parameters easier during flight and postflight analysis. Thus, return on investment is created by faster, more reliable, and cheaper control of flights, complete use of fuel tanks, readjustment of pressure and vibrations, as well as the redesign of rockets for reusability and more efficient and cleaner fuels.<\/p>\n\n\n\n This framework methodology is advantageous for rockets and anything that feeds sensor data and needs to know if it has a fault and where it is located. <\/p>\n\n\n\n It is a general solution that can be used in: <\/p>\n\n\n\n The E<\/strong>uropean iN<\/strong>itiative for L<\/strong>ow cost, I<\/strong>nnovative & G<\/strong>reen H<\/strong>igh T<\/strong>hrust EN<\/strong>gine project is a program<\/a> to advance the trial of cutting-edge production and deployment technologies for reusable rocket engines.<\/p>\n\n\n\n The ENLIGHTEN project and network allows Proekspert AS to continue to apply its advanced technology to the fast-developing space industry with trusted partners. The technology derived is of vital use to machine- and industrial health diagnostics and monitoring.<\/p>\nAlfonso Garcia-Sosa, Data Scientist – Tech Lead at Proekspert<\/em><\/cite><\/blockquote>\n\n\n\n The Enlighten consortium, headed by ArianeGroup<\/a>, comprises 18 partners from eight countries: Austria, Belgium, Estonia, France, Germany, Italy, Netherlands and Portugal. It consists of industrial firms (ArianeGroup GmbH, Avio S.p.A., APP, ADIRA Metal Forming Solutions), research institutes and universities (DLR, Fraunhofer, IREPA, KU Leuven, Onera, Silicon Austria Labs), as well as SMEs and start-ups (Aiko, Areelis, AZO, Edgise, Erneo, Laser Cladding Venture, Proekspert).<\/p>\n\n\n\n ArianeGroup emerged from a joint vision of Airbus and Safran to elevate the European space sector to new heights. Equally held by both entities, it integrates their collective civil and military space launcher activities and expertise into one unified organization.<\/p>\n\n\n\n In July 2022, the ArianeGroup was selected by the European Commission to carry out the Enlighten project, which was given a budget of \u20ac17.4 million under Horizon Europe<\/a>, its framework research program.<\/p>\n\n\n\n\n
\n
What is the ENLIGHTEN project?<\/h2>\n\n\n\n
\n
Who are the ArianeGroup?<\/h2>\n\n\n\n