Artificial Intelligence Across Rail Networks
Loading results...
| Country | Organization | Category | AI Application | Source Link |
|---|---|---|---|---|
| Switzerland | Swiss Federal Railways (SBB) | Asset and Infrastructure Monitoring | AI-enhanced track inspection. SBB employs AI to analyze video from track-inspection cars that run at 120 km/h, in order to detect rail defects. For example, if a crack is found on a rail during one inspection run, the AI can determine if a crack in a subsequent run is the same crack or a new one a few centimeters away. This differentiation, impossible manually at high speed, allows maintenance crews to accurately target new cracks for repair and avoid redundant work, thereby improving infrastructure integrity. | [LINK] |
| USA | BNSF Railway | Asset and Infrastructure Monitoring | Automated wayside sensor analytics. BNSF leverages AI to sift through data from its trackside detectors (over 35 million readings a day from thermal and acoustic sensors, machine-vision cameras, etc.) that monitor train wheel health. The AI identifies signs of issues like overheating brakes or wheel cracks in real time. By learning from these massive inputs, it can predict wheel failures in advance and prompt proactive maintenance, reducing derailment risk and service interruptions. (BNSF"s system monitors 1.5 million wheel instances across its network via this AI approach.) | [LINK] |
| France | SNCF | Asset and Infrastructure Monitoring | AI for escalator and system monitoring. | [LINK] |
| France | SNCF | General Office Work | SNCF Group is implementing generative AI through a dual strategy: broad employee engagement and an internally developed tool. Over 20,000 employees have already been trained in generative AI via workshops, conferences, and e-learning. Simultaneously, SNCF has created its own solution, “SNCF Group GPT,” launched in January 2024. This in-house tool helps employees with daily tasks by delivering fast, accurate responses like ChatGPT, while ensuring strict data confidentiality. The system uses a blend of large language models from providers such as OpenAI, Anthropic, and French company Mistral, chosen for task relevance and energy efficiency. Crucially, SNCF Group GPT operates within a private cloud, safeguarding sensitive data and guaranteeing that neither personal nor corporate data is used to train external AI models. | [LINK] |
| France | SNCF | General Office Work | SNCF Group has also developed an HR chatbot powered by generative AI, which taps into the company’s comprehensive HR documentation database. This tool allows employees to quickly find accurate answers to their HR-related questions, streamlining processes and enhancing overall efficiency within the HR function. | [LINK] |
| France | SNCF | Digital Twin | SNCF Gares&Connexions is pioneering smart transportation infrastructure by partnering with Akila to deploy an AI-powered digital twin platform at Monaco’s Monte-Carlo station. This sophisticated system combines digital twins, AI model training, and real-time AI agents to simulate and manage complex station operations with high fidelity. Powered by NVIDIA Omniverse technologies and the Omniverse Blueprint for Smart City AI, the platform ingests diverse data—including crowd movement, energy usage, and environmental factors—to optimize energy efficiency, improve safety through emergency simulations, and enable predictive maintenance. The result is dynamic, real-time infrastructure management that reduces costs, lowers emissions, and enhances passenger services, setting a blueprint for smart cities and transport hubs worldwide. | [LINK] |
| UK | Network Rail | General Office Work | Network Rail, in collaboration with Oakland, implemented a Generative AI solution to revolutionize its knowledge management by transforming its vast lessons-learned library into an interactive, searchable chatbot experience. Using Retrieval-Augmented Generation (RAG) and Microsoft Azure’s OpenAI services, the AI system enables employees to ask natural language questions and receive precise, cited answers from previously siloed documents, significantly improving accessibility, efficiency, and actionable insights for infrastructure projects. This innovative use case showcases how AI can unlock critical institutional knowledge, streamline workflows, and drive smarter decision-making across large organizations. | [LINK] |
| USA | Union Pacific Railroad | General Office Work | Union Pacific Railroad has embraced artificial intelligence by developing its own secure version of ChatGPT, called UP Chat, to safely harness the benefits of generative AI while avoiding the security and privacy risks posed by public tools. UP Chat allows employees to generate blog posts, summarize white papers, analyze data trends, create project summaries, draft presentations, and even produce code snippets, all within the company’s protected network. Strict policies govern its use, prohibiting external sharing of outputs, automating decisions, or offering legal, medical, or financial advice. Beyond UP Chat, Union Pacific is integrating AI and machine learning across its operations to optimize supply chains, predict maintenance needs, and improve overall efficiency, underscoring its commitment to innovation in delivering premier transportation solutions. | [LINK] |
| Switzerland | Swiss Federal Railways (SBB) | Customer Experience | The Swiss Federal Railways (SBB) have launched “SBB Chat,” an AI chatbot based on ChatGPT to answer customer questions 24/7 about tickets, timetables, and services. Running on Microsoft servers in Europe but storing data exclusively in Switzerland, it ensures privacy by prohibiting Microsoft and OpenAI from using user data for training. The bot handles simple queries reliably—like translations or station facts—but struggles with complex calculations or live timetables, sometimes giving inconsistent fare advice. Overall, it resolves most inquiries without human intervention. | [LINK] |
| Denmark | Danske Statsbaner (DSB) | Safety and Surveillance | Danish rail operator DSB is deploying AI-powered camera systems to detect graffiti on trains and prioritize cleaning, reducing manual inspections and costs. Additionally, DSB is testing AI solutions to enhance safety by identifying unauthorized people on railway tracks. | [LINK] |
| USA | CSX | Customer Experience | CSX has enhanced its ShipCSX online portal by launching “Chessie for ShipCSX,” an AI-powered virtual assistant built with Microsoft Copilot Studio and Azure AI Foundry. Chessie enables customers to interact in natural language to track shipments, manage logistics, and quickly get answers to common questions, improving both speed and convenience compared to manual searches. Within its first 45 days, over 1,000 customers engaged with Chessie in more than 4,000 conversations, signaling strong user adoption and satisfaction with this smarter, streamlined customer service experience. | No link |
| Denmark | Danske Statsbaner (DSB) | General Office Work | DSB Proprietary ChatGPT solution | [LINK] |
| Denmark | Danske Statsbaner (DSB) | Asset and Infrastructure Monitoring | Speech-to-text systems for error reporting | [LINK] |
| USA | CSX | Predictive Maintenance | CSX is transforming rail safety and efficiency with AI and cloud technologies. Using Azure Arc and Azure Local, the company conducts real-time railcar inspections at speeds up to 40 mph, instantly detecting issues like leaky bearings or grain leaks that previously demanded extensive manual effort. Additionally, CSX has rolled out FAA-approved autonomous drone inspections across 13 rail yards, powered by machine vision models from Azure Machine Learning Studio. These drones have cut inspection times by 90% for deployed use cases, boosting yard operations, safety, and inspection frequency, with further plans to expand into beyond visual line of sight capabilities. | [LINK] |
| Germany | Deutsche Bahn (DB) | General Office Work | Deutsche Bahn’s “BahnGPT” is an AI chatbot launched in November, built on Microsoft Azure OpenAI and secured with robust data protection measures. Approximately 8,000 employees use BahnGPT for tasks like document analysis, knowledge sharing, and drafting meeting minutes, helping to streamline workflows and reduce workloads across the organization | [LINK] |
| Germany | Deutsche Bahn (DB) | General Office Work | Deutsche Bahn’s “AuditGPT” is an AI tool designed for the internal audit department, assisting in drafting audit reports. By automating report creation, it frees up employees to focus more on hands-on audit tasks, enhancing efficiency and productivity. | [LINK] |
| USA | CSX | Asset and Infrastructure Monitoring | CSX is enhancing intermodal terminal operations with AI-driven camera technology that automatically detects departing trucks, such as bobtails, at terminal gates without human intervention. Leveraging Azure and real-time AI processing, this solution reduces congestion, speeds up truck turnaround times, and improves driver efficiency, supporting both operational performance and customer satisfaction in the fast-paced intermodal environment. | [LINK] |
| France | Alstom | General Office Work | Alstom has developed an in-house AI tool that supports various business functions, from engineering projects to HR and finance. A key use case involves helping engineering teams write high-quality specifications for railway systems. Using Azure OpenAI services, the tool checks, rewrites, and even generates technical requirements, improving specification quality by 25%. This reduces costly errors, as poorly written requirements can average a full day’s salary in corrective work, leading to significant savings and efficiency gains. | [LINK] |
| Japan | Tokyo Metro | Predictive Maintenance | Tokyo Metro has modernized its subway maintenance by developing an in-house AI solution using Microsoft Azure AI services and the Power Platform. Traditionally dependent on labor-intensive human visual inspections across its 200 km network, the company now analyzes images of railroad tracks to detect equipment deterioration and abnormalities automatically. This innovation aims to improve inspection quality, reduce employee workload, and enhance overall safety, reflecting Tokyo Metro’s commitment to efficient, self-driven technological solutions. | [LINK] |
| Spain | Renfe | Safety and Surveillance | Renfe’s Renfe Smart Security Station (RS3) project is revolutionising security across Spain’s commuter rail network by deploying AI-powered video surveillance systems in nearly 600 stations. Backed by over EUR 32 million from the EU’s NextGenerationEU fund, the initiative integrates more than 9,000 IP cameras and advanced video analytics to detect incidents such as crowding, fare evasion, falls, fires, vandalism, and assaults in real time, enabling faster responses from Renfe’s 24-hour Security Centres. Data from thousands of CCTV feeds is processed anonymously and deleted within milliseconds, ensuring privacy compliance while transforming stations into predictive security environments. Collaborating with innovative tech firms like Infinity Neural and Imotion Analytics, Renfe aims to modernise its infrastructure, boost operational efficiency, and enhance passenger safety and confidence, ultimately supporting Spain’s broader goals for sustainable and attractive public transport. | [LINK] |