Environment-Aware Learning of Smooth GNSS Covariance Dynamics for Autonomous Racing

2026-02-24

AI Summary

This work addresses the challenge of obtaining accurate and temporally smooth localization covariance estimates for high-dynamic autonomous race cars in GNSS-degraded environments, where poor uncertainty quantification compromises control stability. To this end, the authors propose the LACE framework, which— for the first time—models covariance dynamics as an exponentially stable controlled dynamical system. By integrating deep neural networks with attention mechanisms, LACE learns temporal covariance evolution from environmental features, while rigorously enforcing exponential stability and smoothness through spectral constraints grounded in contraction theory. This approach achieves tight coupling between environmental perception and state estimation. Experimental validation on the AV-24 racing platform demonstrates that, under GNSS-degraded conditions, LACE significantly improves localization accuracy, yields more reliable and smoother uncertainty estimates, and thereby enhances the robustness of the downstream control system.

Technology Category

• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Computer Vision: Learning & Optimization for CV

Application Category

• Security and Privacy: Large-scale security measurements
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Ensuring accurate and stable state estimation is a challenging task crucial to safety-critical domains such as high-speed autonomous racing, where measurement uncertainty must be both adaptive to the environment and temporally smooth for control. In this work, we develop a learning-based framework, LACE, capable of directly modeling the temporal dynamics of GNSS measurement covariance. We model the covariance evolution as an exponentially stable dynamical system where a deep neural network (DNN) learns to predict the system's process noise from environmental features through an attention mechanism. By using contraction-based stability and systematically imposing spectral constraints, we formally provide guarantees of exponential stability and smoothness for the resulting covariance dynamics. We validate our approach on an AV-24 autonomous racecar, demonstrating improved localization performance and smoother covariance estimates in challenging, GNSS-degraded environments. Our results highlight the promise of dynamically modeling the perceived uncertainty in state estimation problems that are tightly coupled with control sensitivity.

Problem

• GNSS covariance
• state estimation
• autonomous racing
• temporal smoothness
• environment-aware

Innovation

• GNSS covariance modeling
• temporal dynamics learning
• exponential stability
• attention mechanism
• environment-aware estimation