Session D4: Robust Navigation Using Alternative Navigation Sensors and Solutions

Integrity and Continuity Concepts of a Vision-Integrated Navigation System for a Civil Aircraft During a Precision Approach
Gabriel Thys, Safran Electronics & Defense, and Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse; Christophe Macabiau, Julien Lesouple, Jérémy Vézinet, Anaïs Martineau, Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse; Raphaël Jarraud, Safran Electronics & Defense
Location: Johnson (First Floor)
Date/Time: Thursday, Sep. 11, 4:00 p.m.

Peer Reviewed

This paper investigates the integration of optical sensors (e.g., monocular or stereo cameras, long-wave infrared or short-wave infrared sensors) into the navigation system of a civil aircraft for precision approach operations. The vision-based measurements are tightly fused with a navigation-grade inertial measurement unit (angular drift on the order of 0.01? /h), an SBAS-augmented GNSS receiver, and a barometric altimeter. The vision system relies on landmark-based positioning, where visual landmarks are located on the runway. The primary motivation for incorporating vision is to mitigate GNSS losses of service and possible navigation continuity losses caused by radio-frequency interference. While vision enhances continuity, it introduces additional failure modes, which pose integrity challenges under civil aviation standards. A technical challenge arises: vision integration increases the integrity requirements on the GNSS signals. To address this, a dual-navigation architecture is proposed. It operates with an SBAS-augmented inertial-GNSS fusion as long as GNSS is available, and switches to an inertial-vision fusion when GNSS is declared unavailable. The objectives of this paper are: (i) to formally analyze the impact of vision integration in terms of continuity and integrity, (ii) to propose fault allocation trees for the hybrid system, (iii) to assess vision-specific failure modes, and (iv) to derive integrity and continuity requirements for future vision-based integrity monitoring algorithms.