Influencing travel mode choice

One area which has attracted a lot of attention is the potential for ICT to influence travel mode choice, and in particular to persuade travelers to use either public transport, or slower modes of travel like walking or cycling, rather than to use a car. The rationale is that travelers would make better, more sustainable choices if only they had more information about how to make their journeys in this way, and that better real-time information could allow them to re-plan their journeys as traffic conditions or disruptions required.

There is already some evidence that providing people with information about travel alternatives does persuade them to change their behaviour (as in the pilot Sustainable Travel Demonstration Towns). To date these sort of initiatives involve face to face contact with a travel adviser; an ICT-based approach might help to scale it up.

A key tool in this is the personal travel assistant — a device, service or application that helps the user to plan and execute journeys. Example include an in-car satellite navigation device, or the route planning capability within Google Maps, or the journey planner web applications offered by local and national transport authorities, such as Transport for London’s “Journey Planner”.
PTAs are a definite ‘carrot’ measure. Travelers have a genuine interest in better journey planning and execution. The commercial success of satellite navigation devices (and the popularity of travel planning web sites) is an indication that there is a latent demand for products and services which could do the same job for non-car travelers.

The perfect PTA does not yet exist. Currently available products include:

  • Web-based travel planning tools, including those provided by transport operators and transport authorities (with those provided in Helsinki, Stockholm and Chicago particularly impressive), Google Maps, and specialist products such as the walking route finder
  • Applications on mobile devices, such as the Nokia Maps application bundled with some devices and also the advertising based amAzeGPS
  • Satellite navigation devices – primarily aimed at car drivers but with some cycle-mounted products also available
  • Travel alert services for mobile users, primarily delivered by SMS, also provided by transport operators.

All of these products offer part of what’s needed, but none offer all of it. The opportunity to do develop a full personal travel assistant is still there.

Several technology developments are making it easier to build PTAs. These include:

  • More powerful mobile devices, with better display characteristics more suitable for maps and better user interfaces to enable querying of and interaction with information
  • More powerful mobile networks, which make it possible to access map databases and real-time traffic information
  • The advent of flat rate data tariffs
  • Developments in mobile device software including Rich Internet Application frameworks, which have made it easier for application developers to use web-based tools and content rather than build special mobile equivalents
  • The integration of GPS receivers into mid-range devices, and improvements to the receivers themselves that have reduced the impact on battery life and speeded up time to first fix
  • The widespread deployment of other kinds of positioning technology, including improved network-based triangulation and WiFi location
  • The creation and aggregation of high quality digitised map information and its widespread availability, initially limited to the driveable road network but increasingly extending to walking, cycling and other modes
  • Increased availability of real time traffic information, gathered from increasingly dense networks of sensors and Automatic Number Plate Recognition systems, but also from information gathered from mobile network user data, as in TomTom’s arrangements with Vodafone in the UK and Netherlands

How effective could PTAs be in driving sustainable transport behaviour? Plenty of social science research demonstrates that most journeys are habitual. Users make the same journey in the same way that they normally do, only changing their pattern when forced to do so via a major disruption. Transport researcher Glenn Lyons distinguishes between two models of decision-making – “homo economicus”, seeking the maximum possible information and then making rational choices based on this, and “Homer” (after the cartoon character Homer Simpson), seeking out “good enough” journey plans and then sticking with them.

So PTAs have to offer significant end-user benefits in terms of time, money or comfort, and/or cost little in terms of effort required to learn and use. How users experience this cost will depend in part on the design of the PTA and on how it relates to the rest of the users’ experience, with more technology-savvy users less challenged by the UI.

Supply-side barriers are also important. Perhaps the most important of these, and the one most closely linked to the demand side issues, is the inadequacy of the search paradigm for trip planning.

For in-car navigation systems the task is relatively simple. The user (or the device itself, using a positioning technology such as GPS) inputs a starting point, and defines an end-point. The system either offers a single routing option, or with more sophisticated variants offers a limited choice of routes – the quickest, the non-motorway option, and in the case of the latest release of Garmin software, an ‘eco-friendly’ option.

For non-car inter-modal transport, the task is much more complex. In an urban environment there are likely to several different ways of completing the same journey via public transport – different combinations of metro and bus systems, for example. Different users will have widely differing preferences, with some preferring a route with fewer changes and other one that minimises walking. Of course, this is what search processes are good for, and it’s possible for users to specify preferences as part of the process.

But entering all these parameters raises the ‘cost’ of using Journey Planner, to the point where planning the journey can take almost as long as the trip itself. Failure to enter all the parameters leads to a sub-optimal journey plan.

PTAs are also constrained by the lack of real-time transport information. There are exceptions (notably the excellent systems in Stockholm and Helsinki) which suggest that there is no technical barrier. Moreover, information systems which form part of the travel operators’ own infrastructure and are under their direct control – such as display boards in station and at bus stops, often do have real-time information.

Part of this is due to a lack of generally accepted standards for information transfer. There is no agreed standard for real time transport information, though there are various standards bodies and ad hoc groups working on it.

Another factor is transport operators’ reluctance to part with ‘their’ information, which they believe could at some point have commercial value. So even schedule information is rarely provided in a friendly format to application developers or content aggregators. Where this is available developers have been able to provide some remarkable information sites – for example,, which shows local and long-distance trains moving on a map, with each train tagged with its speed, destination and interim stops, purely on the basis of the schedule.

The data sets used for personal travel assistants are sometimes inadequate for inter-modal travel planning, because they are based on information originally collected and collated for vehicle-oriented satellite navigation devices. For example, a short journey in central London planned via Google Maps ignores the option of walking through the parks, even when the ‘walking’ option is selected; Google attributes this to the data set it uses, which does not include roads not accessible to car drivers. Similarly, there is an ongoing petition asking Google to add a cycling option to the directions on its maps.

User interfaces designed for the PC do not work well on mobile devices, and positioning technologies still do not work well enough to enable the user’s current location to be provisioned by the device itself; this problem is most keenly felt for unfamiliar journeys, which are where personal travel assistants are likely to be most useful. It is worth noting that Nokia’s OVI map service allows users to plan their journey via a PC and then to send the resulting route to their mobile device. This feels like a work-around rather than a proper solution, but at the same time points the way towards applications which are accessed from, rather than resident on, mobile devices.

The most impressive PTA of all is one that does not really exist, but is depicted in a short film made for the City of Amsterdam as part of a Cisco project on ‘connected urban development’. This shows an office worker’s PTA which is linked to the company’s parking reservation system, its carbon accounting system, its hot-desking planner, and the end user’s social network and home automation system. All this is in addition to the interfaces with transport information, walking and cycling routes, weather forecasts, travel booking and payment systems. It can be accessed from multiple devices and public transport kiosks, and supports push alerts as well as queries.

The office worker’s PTA speaks to him in English, but with a rather attractive Dutch accent. Future PTAs are likely to include some aspects of avatar technology in the user interface, particularly if they are to become acceptable to a wider range of users. BT is already working on Journey Angel, a PTA to be launched in London for the 2012 Olympics, which will incorporate some aspects of avatar technology in the user interface; this may be used to provide multi-language support, whereby directions are sent to an application as codes and then delivered to the user as speech in the appropriate language.

The use of intelligent agent software, a branch of artificial intelligence, may also help to support more user-friendly PTAs which could help to get around the problem of multi-dimensional searching. A University of California project as early as 2001-2 attempted to design an application known as the “Travel Elves”, which could perform the heavy search and analysis activities on the basis of pre-defined preferences, finally showing the user a set of options and/or a best possible travel plan. Since then the use of agent technology there have been few examples of practical implementation; but increasing processing power in devices and hosted applications may lead to a fresh look.

Leave a Reply

Your email address will not be published. Required fields are marked *

Loading Facebook Comments ...