Case Study: a major European aircraft manufacturer achieves backup airspeed calculation with Expleo’s deep learning solution

Deep Learning-based application for backup airspeed calculation

Expleo was engaged by a major European aircraft manufacturer to address a critical safety challenge. The customer needed a backup system for calculating airspeed to mitigate the risk of temporary failures in its primary Pitot tube sensors, which can be obstructed by ice crystals or insects and lead to inaccurate speed readings.

Expleo developed a deep learning-based application using TensorFlow and Keras. The solution analyzed data from engine pressure sensors and blade rotation speed through a neural network model to provide accurate backup airspeed calculations. The model achieved a margin of error of only 10 knots and was validated by the customer, promising enhanced robustness and accuracy for flight plans while automatically detecting sensor anomalies. This success further demonstrated Expleo's expertise in aeronautical engineering and data science.