A sustainable shuttle solution

A new mode of transport: ride-sharing on-demand in multi-user vehicles

Urban mobility faces challenges that can only be resolved with ingenious and creative sustainable solutions. Smart Cities need to reduce the emissions and pollutants produced by urban transport and decrease traffic flow and congestion. Solutions need to match what citizens want: safe, reliable, comfortable and eco-friendly transportation, and to avoid traffic jams as much as possible. The third variable in the equation is car manufacturers, which are considering new mobility services aligned with city and user demands by researching opportunities based on the idea of “Mobility as a Service“.

Researchers from Universitat Politècnica de Catalunya (UPC) trying to solve the equation have developed a prototype: a demand responsive shuttle which combines Information and Communication Technologies (ICT) with the concept of ride-sharing.

Sustainable transportation is not just a question of technology. New automotive technology and ICT  apps combined with social changes, are leading to a shift from vehicle ownership to vehicle use. The model enables any user to access transportation services and allows them to travel anywhere, at any time.

 

The prototype

The prototype is a model of a fleet of multiple-user vehicles circulating around the city, all connected to a server that centralises user demands. Technical challenges lie in dispatching the fleet in real time and dynamic route calculation. Multiple passenger ride-sharing is possible through the application of Information and Communication Technologies. Users just have to open their app and specify their location and final destination, as well as their desired departure and arrival times.

To create the prototype, the researchers have analysed and evaluated both the technology required through microscopic simulation emulating real-time traffic information whilst interacting with a Decision Support System. The project addresses how to access the system, manage the fleet, assign services and increase route flexibility, among many other issues.

 

Other on-demand services

Helsinki launched Kutsuplus in 2012, an innovative bus service as part of a long-term plan to make cars redundant. It was one of the world’s first attempts to reinvent carpooling for the age of algorithms.

The service matched passengers who were headed roughly in the same direction with a minibus driver, allowing them to share a ride that cost more than a regular city bus but less than a taxi. Operated by the Helsinki Regional Transport Authority, Kutsuplus was the best-known feature of Helsinki’s and Finland’s intelligent traffic system. However, Helsinki authorities halted Kutsuplus on 31 December 2015, deeming its cost to the taxpayers too high.

Sharing on-demand transport in autonomous vehicles (shared taxis, autonomous shared vehicles, etc.) is a growing area of study and has potentially high impact results. Just to mention two examples: Google is developing its own competitor to Uber, and New Jersey is looking into an autonomous taxi system.

There are already products in the market for this, but through our research we tried to find the most efficient algorithm for the routing.

 

How the goal is achieved?

The main goal of the project was to analyse the proposed transport model in an urban area and its impacts on traffic and transport operations. Large-scale tests were carried out, which included:

• Developing a simulation system able to model user requests and fleet management.
• Solving the vehicle routing problem, taking into account time-dependent traffic dynamics on an urban scale.
• Using microscopic traffic simulation to simulate traffic dynamics.

 

Technical approach

Traffic simulator:

• Realistically emulates the flow of individual vehicles in a road network.
• Produces up-to-date, dynamic traffic conditions (i.e. time-dependent link travel times).
• Transports vehicles along the road network, keeping track of their positions and state.

New developed components:

• Supports dynamic routing and scheduling of the fleet vehicles according to passenger requests.
• Time-dependent routing algorithms, which use real-time traffic conditions provided by the simulator.

Results & Conclusions

A fleet size of 500 vehicles can support a percentage of the demand in Barcelona Eixample with a reasonable ratio of accepted requests:

• 90.7% of accepted requests for 2,600-2,800 passengers (5% of rush hour demand)
• 78.6% of accepted requests for 5,000-5,200 passengers (10% of rush hour demand)

Results analysis show that the new transport model is beneficial for cities in terms of reducing traffic flow and parking space requirements, and is also beneficial for users by reducing their total travel time.

 

Main project outcomes

An architectural model has been built on top of the AIMSUN microscopic traffic simulator to simulate and evaluate the new transportation service.

Further transportation strategies linked to the concept of ride-sharing could be evaluated using the technology developed, by either improving or developing current algorithms.

 

Authors and project researchers

This study was held by PhD researchers and engineers in inLab FIB, part of the Barcelona School of Information Technology at UPC. inLab FIB specialises in research and the innovative investigation of applications and services based on the latest information technology.

It is one of the first projects developed the platform CARNET, and was supported by Volkswagen Group Research and by SEAT’s Innovation and Alternative Mobility department.