Amiral Technologies rejoint Nuclear Valley, le pôle de compétitivité de la filière nucléaire française
”Etre membre du pôle de compétitivité Nuclear Valley est une évidence pour nous pour au moins trois raisons : la filière nucléaire est le domaine
Engines of the same category are produced in high quantity. They all have the same structure and theoretical life time, however, during usage, each engine is used in different contexts and its real remaining useful life is different from the other ones.
Prediction of Remaining of Useful Life for turbofans based on learning data for 218 units (NASA Open Data) in order to alert before failure, reduce estimation errors and optimize maintenance planning.
The model we developed, based on progressive learning from successive indiviual equipment failures, allowed monitoring of a large number of similar engines and provided a reliable prediction of the Remaining Useful Life as it learnt on data collected after every flight cycle.
After 30% of the fleet reaching its end of life point the prediction of the end of life date for the other engines is quite precise and it prediction error reduces as new data arrives.