SmaCS will conceive a camera-based prototype solution, validated in relevant environment in the CleanSky2 Integrated Cabin Demonstrator, for digitalized on-demand verification of TTL - Taxi,Take-off and Landing requirements for cabin luggage. It will be designed to be highly reliable, cost effective (6 seats minimum per camera) and easily upgraded with additional camera-based TTL cabin requirements verification functionalities. To fulfil this ambition, the consortium will offer a disruptive approach based on 3 main pillars: A Machine Learning algorithm, based on Deep Neural Networks (DNNs), for cabin luggage recognition in low light and low contrast environment built from libraries and specific developments; an innovative way to produce learning dataset based on videos coupled with 3D models; an aircraft compliant, ultra-light, ultra-compact Image data processing platform with highly adaptable CVMS interface connection capabilities.
The project is positioned to address the goal of designing and developing a camera-based prototype for digitalized on-demand verification of TTL requirements for cabin luggage.
SmaCS will span cooperation between hardware providers and computer vision experts responding to the increasing importance of smart system inclusion in the aircrafts. The technological part considers the inclusion of reliable perception systems based on deep learning technology to detect and localize the objects of interest, i.e., luggage, cabin components, human body parts, and their visual relationships, covering all potential use cases including partial or total occlusions of the objects with body parts of the passenger.
Otonomy Aviation and Vicomtech set the grounds for the project deployment and discussed with the Topic Manager (SAFRAN) about detailed requirements and needs, in order to start building the path on the activities of the WP3 System Definition and Specification, that will serve as the bottom line of the development of the system components and bricks.