Advanced Driver Assistance System for Trams and Metros

Motivation

Within the complex multi-modal environment of busy urban areas, trams often share roads with a variety of other users, including cars and trucks. Other users, such as pedestrians, cyclists, and motorcyclists, are highly likely to be harmed should a collision occur with a tram. As trams increasingly provide an important part of a more sustainable urban transport solution, the frequency and severity of tram related collisions should not be underestimated.

To ensure optimal safety in this congested environment, tram drivers are required to monitor multiple operationally critical areas simultaneously. These include the area in front of the tram, and along its length, the doors during boarding and alighting of passengers, and known blind spots.

When combined with performing all other required driving activities, these monitoring tasks can become overwhelming and lead to a reduction in overall operational safety. It is no surprise that despite suitable training and resourcing, human errors are the most frequent cause of tram accidents.

However, studies confirm that a 30% reduction in accidents of this type can be achieved by adopting suitably functional Advanced Driver Assistance Systems (ADAS). Our prototype HORA has been designed to deliver against this requirement.

To understand our clients’ needs, Hitachi European R&D conducted an ethnographic and ergonomic study with a diverse set of tram drivers to define required supportive ADAS features and functions. These included enhanced hazard detection across occluded areas.

Development and Implementation

Hitachi European R&D, in co-creation with Hitachi Rail STS, has developed a generalised predictive perception and effective hazard warning system for trams that ensures safer public transport and enhanced network efficiency.

The system has been integrated onto an operational tram operated by Hitachi Rail’s customer ANM (Neapolitan Mobility Company).

Sensor setup on customer’s tram (a) and sensor coverage around the tram (b)

Our prototype HORA ADAS combines a state-of-the-art sensor setup, with an advanced predictive perception algorithm. This enables the tram to detect and track objects continuously, delivering realtime predictions of potential hazards that may be partially occluded and not easily seen from the tram’s cabin.

When required the HORA system alerts the tram driver via its intuitive human machine interface located in the tram’s cabin.

In this way HORA provides a new source of immediately accessible information to support the driver’s decision-making and ensure correct responses to potential dangers.

Testing and Validation in Napoli, Italy

Rigorous testing and validation were conducted to ensure the system’s reliability and effectiveness. This included real-world testing in Napoli, a city with complex traffic flow, blind-spots, and pedestrian crossing scenarios.

With HORA’s state-of-the-art sensor package providing a continuous situational data feed, our advanced predictive perception algorithm accurately predicts emerging hazards, including occluded pedestrians, objects in blind spots, or simultaneous dangers from multiple directions. HORA’s alert functions then effectively convey this critical information to the tram driver.

Experienced drivers at ANM confirm they found our prototype HORA very supportive and a useful new tool in their day-to-day tram operations.