Improving Tram Safety and Efficiency with ARISE

Urban mobility needs an urgent change to meet climate and efficiency goals set by the administration. As the European Union aims for a 90% reduction in transport-related greenhouse gas emissions by 2050 under the European Green Deal, sustainable, low-emission transport modes like trams are taking centre stage to achieve that goal. However, tram networks still face operational inefficiencies, driver shortages, and frequent accidents. The project ARISE seeks to address those challenges.

The project, co-funded by EIT Urban Mobility and led by University College London, brings together academic institutions, public transit agencies, and mobility tech firms to develop and test OTIV.TWO, an AI-based Advanced Driver Assistance System (ADAS) designed for urban trams.

What is OTIV.TWO?

OTIV.TWO is a modular, scalable ADAS that uses artificial intelligence, LiDAR, and camera-based sensors to provide situational awareness in real-time to tram operators. The system processes environmental data and detects potential hazards, improving decision-making.

The technical components of the solution are:

• • • • LiDAR sensors: They provide high-resolution 3D mapping of the surroundings, identifying obstacles with millimetre precision, even in challenging weather or lighting conditions.
      • Cameras: They capture visual data for object classification, traffic signal recognition, and scene segmentation.
      • AI-powered algorithms: With the real-time data gathered through the LiDAR sensors and cameras, they continuously assess the environment, predict the movement of objects (pedestrians, cyclists, vehicles), and issue proactive warnings.
      • Human-Machine Interface (HMI): It displays real-time alerts and recommendations to the driver through visual and audio cues.

Thanks to the use of these components, OTIV.TWO aims to enhance tram operations by offering functionalities such as these ones to operators:

• • • • Collision avoidance: Detects static and dynamic objects in the tram’s path and issues early warnings or interventions to avoid accidents.
      • Traffic signal recognition: Identifies and interprets traffic lights and signage relevant to tram operations.
      • Curve-speed monitoring: Ensures safe navigation of curves by evaluating real-time speed against route parameters.
      • Real-time co-pilot mode: Acts as an intelligent assistant to the driver, offering context-sensitive feedback.

As an additional benefit, the developed system could be adapted to various tram models, including legacy fleets, and does not require structural changes to the vehicles or tracks, making it suitable for widespread adoption.

Performance and Evaluation

As part of the project, an evaluation process to assess the effectiveness of OTIV.TWO across technical, operational, and human factors has been developed. The metrics that could be assessed, but not restricted to, are:

Technical Metrics:

1. 1. 1. 1. Accident Reduction
            • Incident rates before and after installation are compared using statistical methods such as t-tests to determine significance.
            • Metrics include:
              • Serious incidents per million kilometres (HRV index)
              • Medium and minor incidents
              • Change in incident frequency post-installation
         2. System Accuracy & Reliability
            • Error detection rate: Percentage of successful collision prevention
            • False positives/negatives: System’s ability to correctly detect vs. misidentify hazards
            • Environmental resilience: System performance in diverse weather and light conditions
            • Manual override frequency: How often drivers need to disable the system
         3. Operational Impact
            • Improvements in punctuality, reduced delays, fewer unscheduled stops
            • Enhanced vehicle availability and energy efficiency (measured in CO₂ emissions and energy consumption per passenger-km)

Human-Centred Design

Another study accompanies the technical evaluation, aiming to help the team understand how drivers interact with the new system and how it affects their stress and workload. This can include the evaluation of items such as:

• • • • Physiological monitoring (heart rate variability, skin conductance)
      • Eye-tracking and EEG analysis during high-load scenarios
      • NASA-TLX and System Usability Scale (SUS) surveys to capture subjective perceptions

This dual approach that combines objective physiological data with subjective feedback ensures that the system enhances safety without overburdening operators.

European Pilots: Real-World Validation

To validate the technology under diverse conditions, pilot projects are being deployed in:

• • • • Zaragoza (Spain) – A modern, semi-isolated tram network
      • Lisbon (Portugal) – A historic, street-level network with mixed traffic
      • Antwerp (Belgium) – A dense, multimodal system with tunnel and street integration
      • Barcelona and Utrecht will follow as scalability case studies

The pilots aim for a 95% obstacle detection accuracy, a 10% reduction in avoidable incidents, and a 5% improvement in travel times.

Conclusion: A Smarter Future for Urban Transport

The ARISE project showcases how AI and sensor technology can be integrated into traditional public transport systems to create safer, more efficient, and more sustainable urban mobility. OTIV.TWO represents a new generation of tram operation support tools, combining real-time awareness with predictive intelligence.

If successful, ARISE could serve as a blueprint for AI-driven enhancements in public transportation worldwide, supporting cities in their transition toward climate-neutral and smart mobility systems by 2030. ARISE exemplifies European collaboration across sectors: technology (OTIV), academia (UPC, UCL, CARNET-UPC), transport operators (Tranvías de Zaragoza, CARRIS, De Lijn, TRAM Barcelona, and Province of Utrecht), and industry (CAF). This ecosystem, supported by EIT Urban Mobility, is driving the future of safe, smart, and sustainable tram travel.

This project is supported by EIT Urban Mobility, an initiative of the European Institute of Innovation and technology (EIT), a body of the European Union. EIT Urban Mobility acts to accelerate positive change on mobility to make urban spaces more liveable. Learn more:eiturbanmobility.eu