Trapeze Blog

The Future of Intelligent Transport Systems

By Brian Higbee January 11, 2021 Reading time:

In the late 1960s, Trapeze developed one of the world’s first fleet management systems. The idea was simple - collect vehicle information, and then use insights to improve timetables and provide basic operations monitoring in real-time, with control of the fleet via voice radio. The system proved a great success in Zurich, Switzerland, and was expanded to other European cities which helped improve public transport service delivery.

This ethos of supporting public transport operations, by providing tools to deliver reliable services throughout the day, has continued as a focus of Automated Vehicle Management (AVM) systems. AVM systems have been continuously updated to include passenger information, as well as more sophisticated analysis and control mechanisms, which have become the Intelligent Transport Systems (ITS) of today.

While technology has progressed considerably since the 1960s, more transport innovations are about to arrive that will transform the industry even further.

So, what does the future hold for ITS?

Service Predictions

Today, the best predictions of service arrival times at a stop use a combination of current and historic vehicle timing information. This history-based prediction uses a combination of short- and longer-term travel times, averaged over many services on the same routes. It provides a more accurate prediction than an individual vehicle delay – where there are sufficient vehicles on a link to generate a statistically meaningful correlation.

With the advent of data warehouses and machine learning, these algorithms are being extended to look for scenarios that can be used to predict the development of a queue before it happens, which in turn informs the prediction algorithm.

Prediction Algorithms - Intelligent Transport Systems

As it is not possible to predict specific events, such as a breakdown, accident, or fallen tree, these will still impact on the first service to encounter the issue. However, smart algorithms can then be used to inform the prediction model for all later services.

The predictions are also influenced by the class of vehicle. A tram on a segregated track is only impacted by other track users, while buses in a segregated bus lane may, in some parts of the network, still be subjected to external factors. Common factors include passenger numbers and weather conditions. 

Bus Electrification

Many transport authorities are investigating the implementation of electric bus fleets. The benefits of electric buses are significant - not just for air quality improvements, but also the reduced maintenance levels required as electric vehicles have minimal moving parts. While they require sophisticated electronics and management software, as well as charging infrastructure, they provide a better passenger experience - with the managed application of full torque available from a stationary position. The ride is also a lot quieter, as the electric motor generates minimal noise and vibration compared to conventional buses. 

Electric buses, unlike their diesel counterparts, emit no direct pollution. However,

Electric Bus charge times

Early model electric buses did not have a long range. An overnight charge would last approximately 170 kilometres, allowing for a morning run, followed by a midday charge to be ready for an afternoon shift.

Today, via improvements in new bus and battery technology, electric bus range expectations are up to 350 kilometres. For shorter-range buses, charging stations are still required, along with the management of this infrastructure.  In addition, vehicle information and status - particularly charge levels - is required so that operations can be managed and charging times can be planned. 

If a bus misses its recharging slot, this will impact not only its own block of trips, but potentially other bus blocks as well. For longer-range electric buses, charging can take place in the depot, and whilst there is less infrastructure involved, it still needs to be matched to the charge levels of the returning buses to maximise availability and minimise electricity costs.

Electric Bus Recharging Station

Building more charging points is a possible solution, but these very high-capacity chargers (capable of 300 KWh or greater charging levels) are expensive, and their utilisation needs to be managed. Systems that know the bus network, vehicles, and the charging locations, can make accurate predictions on the remaining bus ranges, which in turn provides information on the required charge times. These charging systems will be essential, as the availability of chargers and the allocation of recharging slots need to be fed back into the real-time management and prediction system.

Automation

Machine learning relies on vast datasets, and often requires a high level of human intervention to train the machine - coding up the various conditions and consequences so that the system can learn quickly. With an AVM solution, this need for extensive training datasets can be met from the information already collected. For example, service controller actions and the state of the network is recorded, both before and after actions are already known, and the system can learn from the consequences.

This offers the possibility to develop learning algorithms that extract from historic information and align the dispatcher interventions that work. These algorithms can then analyse the transport network, looking for signs of a potential incident.

Intelligent Transport Systems Disruptions

At all times, the system should keep the service controller informed of the situation on the road, and automation should assist in taking actions without the service controller losing any control.

These learning tools are coming, building on the large store of data already available in existing systems, as well as collecting new information as these systems grow and evolve.

Conclusion

With the growing availability of sensors and monitoring devices - all part of the Internet of Things trend - larger datasets will flow into the control centre, and the tools used to monitor and manage this data will flourish. Transportation systems can benefit from these general trends, introducing more cost-effective enhancements, in the same way the smartphone has revolutionised the distribution of passenger information.

The future is almost here!

For more information in Intelligent Transport Systems, visit the Trapeze ITS Hub.

Author: Brian Higbee

Brian is a Project Manager and System Engineer with more than with 22 years of experience in implementing AVLC and AVM solutions for major clients globally, including Transport for London. Brian’s expertise also extends to embedded mobile solutions, data and voice communications, real time data processing, central control systems and database solutions. Motivated by bringing together complex multi-disciplinary engineering solutions, Brian thrives on delivering solutions to time and within budget. Connect with Brian on Linkedin.

Contact Us Schedule a Demo View Solutions