[Report] Tram Stops When Bus Does... The Future of Railways with AI at the Controls [InnoTrans 2026]

Blueprint for Autonomous Hydrogen-Electric Trams

AI Handles Obstacles and Emergencies, Optimizes Dispatch and Maintenance

Fully Autonomous Trams Targeted by 2031 as Technology Accelerates

While the tram was running through the city center, a bus suddenly stopped in front of the tracks. The tram reduced its speed to maintain a safe distance and came to a halt. A short while later, as the bus cleared the way, the tram started moving again. There was no driver at the controls. It was a decision made autonomously by artificial intelligence (AI), which had assessed the surrounding situation.

On the 22nd (local time), visitors at the Hyundai Rotem exhibition hall at InnoTrans 2026 in Berlin, Germany, were examining the Artificial Intelligence (AI) autonomous tram system. Photo by Seomideum.

On September 22, 2026 (local time), at Hyundai Rotem’s exhibition hall during InnoTrans 2026, the world’s largest railway trade fair held in Berlin, Germany, the future of railways was on full display in a model city set up at one side of the booth. On the meticulously crafted diorama, packed with buildings, roads, stations, and depots, the tram ran its route while a large screen on the wall showed the vehicle's position and operating status in real time. The words "AI Autonomous Tram System" greeted visitors.

Visitors stopped in front of the diorama, took out their phones, and snapped photos of both the screen and the model. At this year’s InnoTrans, Hyundai Rotem put the spotlight on its fully autonomous railway that integrates AI into a hydrogen-electric tram.

The working day for a future tram begins with AI even before operations start. The AI checks the previous day's operation records, vehicle status, and hydrogen fuel cell charge levels. Using a "digital twin"—a virtual replica of reality—AI factors in expected passenger numbers and urban traffic conditions to simulate thousands of possible driving scenarios in advance. Any vehicle found to have an issue is pulled from service, and the system calculates the optimal dispatch intervals and travel times between stations.

On the 22nd (local time), an official at Hyundai Rotem's exhibition hall at InnoTrans 2026 in Berlin, Germany, explained the advanced driver assistance system (ADAS). The screen displayed camera footage alongside surrounding information recognized by LiDAR. Photo by Seomdeum.

During actual operation, LiDAR and cameras serve as the operator's eyes. On the Advanced Driver Assistance System (ADAS) screen installed on the wall of the Hyundai Rotem booth, images captured by the cameras and surrounding objects detected by LiDAR appeared simultaneously. The AI does more than simply check if there is an object ahead; it distinguishes between people and structures, and even judges the likelihood of an object entering the tram's path.

Unlike cars, trams cannot swerve to avoid obstacles. Since roads are shared by vehicles, pedestrians, and bicycles—and trams cross intersections and pedestrian crossings—the ability to detect danger early and stop in time is extremely important. This is why Hyundai Rotem believes the level of autonomous driving technology required for trams is higher than that for subways.

On the 22nd (local time), a model of a hydrogen electric tram was displayed at Hyundai Rotem's exhibition hall at InnoTrans 2026 in Berlin, Germany. Hyundai Rotem aims to develop fully autonomous driving technology by 2031 by combining AI-based autonomous driving technology with hydrogen electric trams. Photo by Seomideum

The role of AI does not stop at driving. In the model city, if a passenger collapses onboard, the AI-powered onboard CCTV system detects the emergency. A warning immediately appears on the control screen, and an ambulance is dispatched. The system simulates a scenario where the ambulance is sent ahead to the station where the tram will arrive, minimizing patient transfer time.

Walking further along the exhibition hall wall, visitors encounter displays showing "CBM (Condition Based Maintenance)", "Digital Factory", and "MMIS+ (Maintenance Management Information System)". The idea is for AI to take charge of not only 'driving' trains but also 'manufacturing and repair' processes.

CBM analyzes vehicle data collected through cameras and sensors to predict the level of component wear and the optimal maintenance schedule. Currently, CBM systems that collect data from sensors installed on key parts are already in use on KTX-Cheongryong trains. Hyundai Rotem explains that, going forward, if the AI is trained on data accumulated from operations, it will be able to predict potential failures before they occur.

The adjacent Digital Factory screen displayed various stages, including "manufacturing simulation," "ergonomic analysis," "process validation," and "factory flow simulation." This showcases the concept of running the factory virtually before actual production. It enables error reduction between design and real-world manufacturing, and pre-validates the layout of workers and production processes.

On the 22nd (local time), digital factory technology was showcased at Hyundai Rotem's exhibition hall at InnoTrans 2026 in Berlin, Germany. The screen displayed a simulation analyzing the worker's posture and work environment in a virtual space. Photo by Seomideum

AI is also being integrated into maintenance. MMIS+ is a system whereby, if an issue arises with a vehicle, a technician can speak into a tablet, and the AI analyzes past maintenance records to suggest solutions. While previous systems focused on storing and managing data, the current stage expands to AI making decisions based on accumulated data.

Hyundai Rotem's ultimate vision is a fully autonomous hydrogen-electric tram running through the city with no operator onboard. After conducting a hydrogen-electric tram demonstration project in 2023, Hyundai Rotem began developing its own ADAS system in earnest and completed the development of the railway vehicle ADAS this year. The company is now working on integrating ADAS with the control system and aims to develop fully autonomous tram technology by 2031.

Competitors are also accelerating the development of tram automation technologies. Alstom has developed a system that detects pedestrians and obstacles using cameras, while Siemens is pursuing automation that enables trams to move autonomously within depots for washing and servicing without a driver. Skoda has also developed obstacle detection and emergency braking technologies using composite sensors. Hyundai Rotem is pursuing differentiation by combining these advancements with hydrogen energy. The hydrogen-electric tram uses both hydrogen fuel cells and batteries, allowing it to travel over 150 kilometers on a single charge and operate without overhead catenary lines.

On the 22nd (local time), at the Hyundai Rotem exhibition hall during InnoTrans 2026 in Berlin, Germany, visitors experienced virtual engineering technology using virtual reality (VR) equipment. Hyundai Rotem showcased technology that pre-validates vehicle design and work environments in a virtual space. Photo by Seomideum

On-site, visitors expressed a mix of anticipation and caution. Fariborz Langeroudi, a railway engineer from Canada, commented on AI-powered railway safety technology, saying, "Technology inevitably moves us into the future," and, "We need to trust and embrace it." Having previously worked with Hyundai Rotem on technology transfer projects, he added, "The company’s approach to technology development and customer communication leaves a strong impression."

Giulio Sambugaro, a railway industry representative from Italy, expressed optimism about the prospects for AI-based autonomous trams: "While there may currently be some caution regarding safety, as the technology advances, it will become increasingly safer in the near future," he said.

The tram in the diorama returned to its depot. Although operations were complete, the AI’s work was not finished. Cameras inspected the vehicle’s body and undercarriage, and the AI analyzed the findings to formulate the next maintenance plan. From pre-operation inspections and driving to emergency response and maintenance, Hyundai Rotem’s exhibition demonstrated a future for railways in which humans have relinquished the controls—but AI’s decision-making is even more deeply embedded throughout the system.

This content was produced with the assistance of AI translation services.