Innovation
Optimising networks by making ticketing data talk

Where should a bus stop be located to boost ridership? What transport services does a new district need? Xerox’s research teams and ACS Public Transport’s experts are pooling ticketing data, demographic information and behavioural analyses to simulate the impact of changes in public transport infrastructure.

Barely a year after merging, the teams at ACS Public Transport and Xerox’s research centre in Grenoble (France) have just presented their first projects. One basic application offers a means of viewing where and how many validations were made on the public transport network during a given period on a digital map of the city. The user can zoom in on a specific sector and select the time period required. A network operator can hence see whether diverting a line has shifted validations to other stations, or even reduced ridership. "It is a highly pertinent presentation method for monitoring service quality," explains Pascal Roux, ACS Public Transport marketing manager.

Simulation
Simply combining ticketing data with a geographic information system hence produces meaningful results. But the Xerox and ACS research teams are going beyond data visualization. They have already laid the groundwork for the tools to model multi-modal transportation environments, for example to simulate the efficiency of a public transportation network as an alternative to individual modes of transport.

On the basis of travel time alone, the models can already be used to infer basic origin-destination information. The results are shown as areas plotted around departure and destination stations. "But that is not enough to be sure that, given changes, a user won’t prefer to drive in order to save time", points out Marco Bressan, head of the Services Innovation Laboratory at Xerox’s research centre in Grenoble. So even more detail is needed...

Behaviour
Numerous factors influence users’ choices: age, total journey time, quality of travel, price of trip, proximity of stations, purpose of journey, connections, parking availability, other available means of transport, etc. To take them all into account, Xerox’s researchers start with the ticketing data, which they analyze to define typical profiles incorporating the city’s geography and demographics. User surveys provide valuable additional details. All this information is combined to build behavioural usage models that can be used to deliver seamless, sustainable and safe mobility.

Enriched with these new factors, the simulation becomes even more realistic. It can hence be used to test various locations for a new station or line, to optimise the network configuration and ensure that it offers the best alternative to travelling by private car, making it a valuable aid to decision-making.