Session A5: Sensor-Fusion for GNSS-Challenged Navigation

A High Availability Inertial-Vision Data Fusion Using an ES-KF for a Civil Aircraft During a Precision Approach in a GNSS-Challenged Environment
Gabriel Thys, Safran Electronics & Defense, and Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse; Christophe Macabiau, Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse; Raphaël Jarraud, Safran Electronics & Defense; Anaïs Martineau, Julien Lesouple, and Jeremy Vezinet, Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse
Location: Beacon A

In civil aviation, the development of LPV (Localizer Performance with Vertical Guidance) precision approaches, facilitated by the emergence of SBAS such as WAAS and EGNOS, is complementing the traditional ILS approaches. However, this development is disturbed by the increasing proliferation of RF interference (RFI) to GNSS, such as jamming and spoofing. As a result, the approach environment for airliners in the near future may be considered GNSS-challenged as GNSS might no longer be relied upon even in environments where states are making best efforts to secure the RF environment.
While navigation-grade IMUs ensure reliable aircraft's orientation for control needs, they cannot be the sole means to provide an aircraft guidance within the stringent tolerable error bounds of LPV approaches during extended coasting periods. In this article, the drift of the IMU is corrected by registering known position visual landmarks using a monocular camera. Specifically, the navigation camera unit measures the line of sight of the aircraft relative to known landmarks on the runway during the final phase of the approach. The camera is a sensor not depending on external RF sources using absolute reference points that complements well the IMU. This sensor is also complementary to GNSS in terms of availability. At high altitudes, RF signals are clear, the probability of jamming is lower, and GNSS multipath effects are minimal. However, there is no camera observation available. On the other hand, at low altitudes or in urban areas, GNSS signals have a higher chance of being combined by RFI while the position observability using a camera is significantly stronger. This highlights the advantage of incorporating camera-based vision information in situations where GNSS availability is limited, enabling more reliable and accurate navigation. By leveraging the strengths of both sensors, the hybridization approach can provide robust navigation capabilities across a wide range of operating conditions.
The objectives of this proposed paper are as follows:
- Underline the utility of vision-based hybridization in the context of LPV precision approaches.
- Highlight the significance of the chosen architecture to achieve the desired performance level in the aimed context. Emphasis will be placed on the specific architectural choices made to ensure optimal integration and fusion of the IMU and camera data.
- Present the initial findings and results obtained from the proposed hybridization architecture.
In this paper, the elements presented are defined as the following.
First of all, targeted operational scenario is presented: an ultimate LPV200 precision approach for an aircraft in a GNSS-challenged environment, equipped with a high-precision navigation-grade IMU (angular drift of the order of 0.01 degree per hour) and a monocular camera positioned at the front of the aircraft, observing the landing runway.
Then, a specific IMU-vision hybridization approach is proposed in order to address the challenges posed by the limited number of landmarks and the high correlations on the measurement errors elaborated from them, where two points on the runway may be indiscernible from the perspective of the camera, and where distance estimate may be challenging. That is why the quality and correlation of visual landmarks need to be addressed.
This specific hybridization approach has been designed to meet the constraints and requirements of the targeted operation and operational scenario.
Finally, this hybridization architecture is tested in a comprehensive simulator. These tests aim to illustrate the advantages and the accuracy characteristics of IMU-vision-based guidance in successfully guiding an aircraft down to the decision height of the approach. The focus of this study is on the application of inertia-vision hybridization in commercial airliners. Therefore, a detailed analysis of the civil aviation regulation framework pertaining to this application was conducted. The operational requirements were specifically conditioned by this targeted use case, which led to the utilization of specific sensors. This had a significant impact on various aspects of the system, including architecture and data fusion techniques. As a result, this paper introduces a fresh perspective on inertia-vision hybridization by considering a new operational scenario, perfectly matching commercial aircraft guidance and control needs, which necessitates different approaches and solutions to address navigation challenges.
This work represents the first stage of the integrity study on the inertia-vision hybridization applied to civil aviation.