Huawei?Pangu Railway Model leverages AI to automate railway fault-identification processes

Huawei has introduced its new TFDS Solution that aims to transform railway operations with AI-enabled fault identification technology.

The technology has the potential to reduce human error as well as substitute legacy manual processes. When fault identification happens in an old-fashioned way, human railway inspectors have to view thousands of images taken by Train Freight Detection Systems (TFDS) to identify malfunctioning.

The sensor-based TFDS calculates the train’s speed by leveraging the magnetic steel sensors on its wheels. It automatically makes the snapshots every few milliseconds and uploads them to a central server for fault detection.

With the help of artificial intelligence (AI), the fault screening process in the railway industry may become smarter and more efficient. Huawei’s TFDS solution, trained by Pangu Model, is able to locate components, determine component status and warn inspectors of faults or anomalies based on its findings.

The firm claims the new solution can accurately identify various types of faults, first analyzing large components, and then proceeding to identification of local faults. Since false alarms can easily be caused by distortion, blockage, and small size of components, the Huawei TFDS solution has an adaptive enhanced detection algorithm, image reconstruction, and adaptive image fusion to be more precise.

According to the press release, the new solution automates the analysis of millions of captured images. Therefore, ?the number of manually-viewed images reduces from around 4,000 down to 200, allowing the labour force to focus on the remaining 5% of images requiring further analysis. Huawei claims that the detection process for a train can be completed within 8 minutes, while the overall fault identification rate can be up to 99.3%.

Earlier this month, it was reported that AI found its use in shipping, helping to reduce the burden on supply chains, finalize contacts and prevent interaction with firms involved in human rights violations or environment-harming practices.