Hitachi Rail, a global transportation company that powers rail systems in more than 50 countries, is integrating NVIDIA AI technology to help rail operators reduce maintenance costs, reduce train idle time, and improve reliability for passenger transportation.
The company will use NVIDIA IGX, an industrial-grade, enterprise-level platform that delivers high-bandwidth sensor processing, powerful AI computing, functional safety features and enterprise security, for its new HMAX platform to process sensor and camera data in real time.
The HMAX platform eliminates lag time between data collection and analysis, enabling Hitachi Railway customers to more quickly detect tracks in need of repairs, monitor deterioration of overhead power lines, and assess the health of trains and signaling equipment.
Hitachi Rail estimates that preventive maintenance costs about one-seventh of the cost of emergency repairs after unexpected infrastructure failures. The company’s existing AI monitoring system has already reduced service delays by up to 20%, train maintenance costs by up to 15%, and energy consumption by reducing fuel costs at train depots by up to 40%.
The company aims to increase these savings even further with real-time analytics using NVIDIA IGX and the NVIDIA Holoscan platform for sensor processing.
“With traditional digital monitoring systems, it would take days to process data and spot issues that needed our attention,” said Koji Wagatsuma, executive officer and chief technology officer for rail vehicles at Hitachi Rail. “Instead, if we could use NVIDIA technology to make real-time predictions, we could avoid service interruptions and significantly improve safety, reliability and operational costs.”
NVIDIA IGX Powers Real-Time AI Engine
Building on its existing HMAX application collection currently running on data from 2,000 trains and 8,000 cars, Hitachi Rail used NVIDIA IGX and the NVIDIA AI Enterprise software platform to create a new accelerated AI application that helps dispatchers monitor train vehicles and infrastructure. NVIDIA AI Enterprise provides tools, pre-trained models and application frameworks to streamline the development and deployment of production-level AI applications.
These applications will soon be available through the HMAX platform and can be used by the company’s international customer base to process the vast amounts of data streaming from train-mounted sensors, either captured from existing systems or imported from third-party software customers are already using.
For example, in the UK, each Hitachi Rail train is equipped with sensors that report around 50,000 data points every fifth of a second. The AI infrastructure that serves this data flow can send timely alerts to train operators when train or track components need maintenance. AI insights can also be accessed through a chatbot interface, allowing operators to easily identify trends and opportunities to optimize things like maintenance schedules.
“If potential issues are not identified and resolved quickly, they could result in service interruptions, causing significant economic losses for our customers and impacting passengers who rely on these transit lines,” Agatsuma said. “NVIDIA’s AI infrastructure means that for the first time, thousands of miles of rail can receive instant alerts, which we hope will reduce delays and disruptions to passenger travel.”
Bringing benefits in the future
The opportunities go beyond just monitoring trains and tracks.
Hitachi Rail is installing cameras on top of its trains to monitor overhead power lines, identify degraded cables, and prevent critical failures. Traditionally, it would take up to 10 days to process one day’s worth of video data collected by a train. NVIDIA’s fast sensor processing allows data to be processed in real time at the edge, sending only relevant information back to operational control centers for analysis and response.
Learn more about the NVIDIA IGX platform.
Main image provided by Hitachi Rail.
