Smart Railway TFDS Solution

A conventional trouble of moving freight car detection sytem(TFDS) uses trackside high-speed cameras to capture images of freight car components. Images are then analyzed by TFDS inspectors, one by one, to identify potential train faults and generate alarms accordingly.

But, with hundreds of millions of images generated every day, enormous pressure is placed on inspectors, resulting in all too frequently missed faults and alarms that are never raised. That's a big problem. So Artificial Intelligence (AI) and similarly advanced technologies are urgently needed to improve the intelligence of TFDSs, accelerate fault analysis, and relieve the pressure placed on inspectors.

Huawei's Smart Railway TFDS Solution deploys the Huawei Pangu Railway Model as the foundation for pre-trained models, integrating AI into the entire TFDS identification process. This effectively filters out over 95% of fault-free images, to precisely identify over 430 common faults on 67 types of freight cars,the comprehensive fault identification rate is greater than 99.3%. Greatly improve operation efficiency.