<p>This paper presents a robust filtering approach designed for multi-rate and asynchronous multi-sensor systems experiencing non-Gaussian measurement errors. The proposed methodology is built upon the Huber estimator, which combines ℓ₁- and ℓ₂-norm regression techniques, and extends the asynchronous fusion filter (AFF) framework. The asynchronous fusion problem is addressed for a generalized case involving an arbitrary number of sensors operating at varying sampling rates. The Huber-based AFF (HAFF) introduces a novel strategy by correlating process noise with augmented measurement noises, thereby enhancing robustness against impulsive and mixed-Gaussian errors. To validate the performance, the HAFF is applied to a tracking problem involving an atmospheric entry vehicle using radar measurements. Simulations consider diverse scenarios, including synchronous and asynchronous sensor setups, Gaussian and non-Gaussian errors, and varying levels of measurement noise. Comparative analysis demonstrates that HAFF consistently outperforms traditional extended Kalman filter (EKF) and AFF in terms of estimation accuracy and resilience to noise. These results underscore the HAFF’s capability to robustly and efficiently fuse data from multi-rate asynchronous sensors in real-world engineering applications.</p>