Predictive Uncertainty for Runtime Assurance of a Real-Time Computer Vision-Based Landing System

2025-08-13

AI Summary

Addressing robustness and reliability challenges in certifying real-time vision-based landing systems for aviation safety-critical applications, this paper proposes a pose estimation framework with runtime assurance capabilities. Methodologically: (1) it introduces a spatial Soft-Argmax-based probabilistic keypoint regression architecture compatible with multiple lightweight visual backbones; (2) it formulates a principled loss function enabling pixel-level, calibrated uncertainty estimation; and (3) it integrates a residual-type Receiver Autonomous Integrity Monitoring (RAIM) mechanism to enable online output integrity monitoring and anomaly rejection. Experiments on a runway imagery dataset demonstrate that the method surpasses baseline approaches in pose estimation accuracy, achieves sub-pixel–calibrated uncertainty estimates, and significantly improves downstream fault detection reliability. Collectively, this work provides a verifiable, certification-oriented technical pathway for data-driven vision systems seeking airworthiness approval.

Technology Category

• Intelligent Robots: State Estimation
• Computer Vision: Biometrics, Face, Gesture & Pose
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Large-scale security measurements

Abstract

Recent advances in data-driven computer vision have enabled robust autonomous navigation capabilities for civil aviation, including automated landing and runway detection. However, ensuring that these systems meet the robustness and safety requirements for aviation applications remains a major challenge. In this work, we present a practical vision-based pipeline for aircraft pose estimation from runway images that represents a step toward the ability to certify these systems for use in safety-critical aviation applications. Our approach features three key innovations: (i) an efficient, flexible neural architecture based on a spatial Soft Argmax operator for probabilistic keypoint regression, supporting diverse vision backbones with real-time inference; (ii) a principled loss function producing calibrated predictive uncertainties, which are evaluated via sharpness and calibration metrics; and (iii) an adaptation of Residual-based Receiver Autonomous Integrity Monitoring (RAIM), enabling runtime detection and rejection of faulty model outputs. We implement and evaluate our pose estimation pipeline on a dataset of runway images. We show that our model outperforms baseline architectures in terms of accuracy while also producing well-calibrated uncertainty estimates with sub-pixel precision that can be used downstream for fault detection.

Problem

• Ensuring safety and robustness in vision-based aviation landing systems
• Developing real-time, certifiable aircraft pose estimation from runway images
• Producing calibrated predictive uncertainties for fault detection in navigation

Innovation

• Spatial Soft Argmax for probabilistic keypoint regression
• Principled loss function for calibrated uncertainties
• Residual-based RAIM for runtime fault detection