USPTO Patent Grant for Autonomous Vehicle Anomaly Detection

ChangeBridge: Patent Grants - AI & Computing (G06N)

Published March 24th, 2026

Detected March 25th, 2026

Summary

The USPTO has granted patent US12586478B2 to Wing Aviation LLC for an unsupervised anomaly detection system for autonomous vehicles. The patent covers techniques using machine learning models trained on time series data to identify and respond to anomalies, enhancing the safety and reliability of autonomous systems.

View original document View source feed page

What changed

The United States Patent and Trademark Office (USPTO) has granted patent US12586478B2 to Wing Aviation LLC. This patent covers an unsupervised machine learning approach for detecting anomalies in time series data, specifically applicable to autonomous vehicles. The system is designed to process large amounts of untagged data, alternating between fitting and trimming optimizations, to identify anomalies with high accuracy. Upon detection, the system can transmit commands to the monitored system, such as an autonomous vehicle, to respond.

While this is a patent grant and not a direct regulatory mandate, it signifies technological advancements in AI-driven anomaly detection for autonomous systems. Companies developing or operating autonomous vehicles, particularly in the aviation sector, should be aware of this patented technology. The patent may influence future development and deployment of similar systems, potentially leading to industry standards or future regulatory considerations regarding AI safety and reliability in autonomous operations.

Source document (simplified)

← USPTO Patent Grants

Unsupervised anomaly detection for autonomous vehicles

Grant US12586478B2 Kind: B2 Mar 24, 2026

Assignee

Wing Aviation LLC

Inventors

Vikas Sindhwani, Hakim Sidahmed, Krzysztof Choromanski, Brandon L. Jones

Abstract

In some embodiments, techniques are provided for analyzing time series data to detect anomalies. In some embodiments, the time series data is processed using a machine learning model. In some embodiments, the machine learning model is trained in an unsupervised manner on large amounts of previous time series data, thus allowing highly accurate models to be created from novel data. In some embodiments, training of the machine learning model alternates between a fitting optimization and a trimming optimization to allow large amounts of training data that includes untagged anomalous records to be processed. Because a machine learning model is used, anomalies can be detected within complex systems, including but not limited to autonomous vehicles such as unmanned aerial vehicles. When anomalies are detected, commands can be transmitted to the monitored system (such as an autonomous vehicle) to respond to the anomaly.

CPC Classifications

G08G 5/55 G08G 5/57 G08G 5/58 G08G 5/21 G08G 5/53 G05D 1/0011 G05D 1/0055 G05D 1/105 B64U 2201/20 B64U 10/25 B64U 50/13 G06F 2201/81 G06F 11/0736 G06F 11/0751 G06F 11/0772 G06F 11/079 G06F 11/3013 G06F 11/3447 G07C 5/0808 G06N 20/00 B64F 5/60