The Assistant Project

Creating a Smartphone App to Assist Older People When Travelling by Public

Transport

Philip Barham

, Stefan Carmien

and Ainara Garzo

Transport & Travel Research Ltd, 15 Market Street, Lichfield, U.K.

Health Division, TECNALIA, Mikeletegi Pasealekua 1-3, San Sebastián, Spain

Keywords: Personal Mobility, Older People, Smartphone Technology, Public Transport, Personal Navigation System,

Ambient Assisted Living Joint Programme (AAL JP).

Abstract: This is a Position Paper, in the form of a short report on progress of the three-year ASSISTANT project,

which is funded jointly by the European Commission’s Ambient Assisted Living Joint Programme (AAL

JP) and the national funding agencies of Austria, Finland, France, Spain and the UK. ASSISTANT’s

objective is to create a system that older people can use, on a smartphone, to help them to travel

independently by public transport, in comfort and with confidence. The target group for the commercial

product that will be developed by the project is more mobile older people. However, because the app will be

highly personalised, and provide the user with the facility of audible and haptic, as well as visual, feedback,

it should also be appropriate for use by people who are partially sighted, people who are hard of hearing,

and people who have difficulty with oral communication. This paper describes progress made during the

first two and a half years of the project’s life, and outlines some of the ways in which the ASSISTANT app

will be tailored to meet the needs of a variety of users, in a variety of public transport contexts.

1 INTRODUCTION

The ASSISTANT Project’s objective is to produce

an effective, and commercially viable, system which

provides accessible support for older people when

using public transport. This app will provide an

online facility for trip planning, guidance during

multi-step journeys on public transport vehicles –

which will include assistance with transfers from

one vehicle to another, and between different means

of transport – and navigational guidance from the

user’s last stop on the public transport network, to

his or her final destination.

As such, ASSISTANT addresses common,

everyday concerns that many older people have with

not being able to find the correct stop, with not

knowing which vehicle to board, with failing to get

off the vehicle at the correct location, and,

ultimately, with getting lost. The app will seek to

bridge this confidence gap during the planning and

making of a journey by public transport, and will

also provide a personalised “safety net” feature that

will enable the user to readily summon help from a

relative or carer of his or her choice, if necessary.

Ultimately, the contribution made by ASSISTANT

will be to encourage the mobility, and thus the social

participation, of older people in Europe, enabling

them to freely access important goods and services,

and fulfil their social and cultural needs, using more

sustainable means of transport. ASSISTANT was

developed with the goal of providing a system for

smaller transport agencies that do not have the

resources to create their own.

The main target group of the ASSISTANT

project is more mobile older people, particularly

those travelling to unfamiliar places or who might,

say, be using public transport after losing the ability

to drive. The decision to concentrate the project’s

design efforts on this sector of the population was

taken partly for commercial reasons, to ensure that

the project’s end-product would be attractive to as

large a market as possible, but also to facilitate the

use of public transport for as many people as

possible. However, because the app will be highly

personalised, and provide the user with the facility

of audible and haptic, as well as visual, feedback, it

should also be appropriate for use by people who are

partially sighted, people who are hard of hearing,

253

Barham P., Carmien S. and Garzo A..

The Assistant Project - Creating a Smartphone App to Assist Older People when Travelling by Public Transport.

DOI: 10.5220/0005434902530258

In Proceedings of the 1st International Conference on Information and Communication Technologies for Ageing Well and e-Health (ICT4AgeingWell-

2015), pages 253-258

ISBN: 978-989-758-102-1

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

and people who have difficulty with oral

communication. This high level of accessibility is

made possible by the fact that smartphones already

have the functionality of allowing people with a

range of sensory impairments to customise their

‘phone to suit their particular needs. If, for example,

the smartphone user has a screen reading device,

designed for blind and partially-sighted people, that

enables him or her to read web pages using the

smartphone, then the device will enable this user to

interact with the ASSISTANT app in the same way.

An important aspect of the work of this three-

year project has been the on-going process of

“horizon scanning” for new apps and devices having

similar functions and objectives to those of

ASSISTANT. This was to continually position the

project’s app in the market.

For example, for the iPhone, the HopStop app

provides directions for the user when making

journeys by public transport, on foot and by bicycle,

but this service is restricted to specific cities.

Although the app can be used in some 100 cities,

including London and Paris, the majority of these

cities are in North America.

Android users can use the SMART-WAY app

for planning trips by public transport. This app also

derives information form the GPS capability of the

user’s ‘phone in order to ascertain the user’s current

position, which is needed in case of having to plan

an alternative route to a destination, or if the

destination is changed. Android is also responsible

for GeorgiePhone, which is a family of apps

specifically aimed at older people and blind and

partially-sighted people. Georgie is able to give

information on the user’s location and the direction

in which he or she is travelling, as well as advice on

which public transport vehicle to board, and when to

get off. Similarly, the RATP, which oversees public

transport in Paris, has developed an app for the

city’s Métro system, which advises the user on

which vehicle to board, and on which door to alight

from in order to minimise walking distance.

Some public transport-related apps have been

developed by individual cities and transport

operators. For example, Traveline Scotland provides

a free app which enables users to plan trips by any

mode of public transport, anywhere in Scotland.

In technical terms, a Unique Selling Point (USP)

of the ASSISTANT system will be the high degree

to which the product will be able to be personalised

according to the user’s ability levels. Information

provided by the app will also be contextualised

according to the situation the user is in, during any

part of the journey. A further important feature of

the product will be personalisation through its error

trapping and remediation functionality, since it will

be designed to be responsive to the user’s physical,

cognitive and mental capacities and preferences.

2 TECHNICAL

CHARACTERISTICS OF THE

ASSISTANT APP

2.1 Hardware and Platforms Required

Whilst the design of the ASSISTANT app will have

innovation at its core, its use will be designed for,

and based on, tried and trusted devices, namely the

home Personal Computer and the smartphone. This

provides the system with a robust and reliable basis.

The user’s smartphone will be the medium through

which information is presented, and the application

running on a PC will enable route design.

ASSISTANT will be helpful even on routes that are

familiar to the user, since the haptic mode reminder

of arrival at the desired destination or transfer point

can allow the user to read or relax when travelling,

removing the need to constantly check the progress

of the journey.

ASSISTANT is a web service based application

that will function by retrieving continually updated

data in the form of maps and schedules. This will be

part of a three-tier system architecture, consisting of

a web based route design interface, a web server

layer implementing the application and a database of

system data (see Figure 1). User-specific data will

also be stored on the smartphone. The route editor

interface will enable users to update their profile

data, and to create and edit a list of favourite

locations to visit, and design pathways to them.

Figure 1: ASSISTANT system architecture.

ICT4AgeingWell2015-InternationalConferenceonInformationandCommunicationTechnologiesforAgeingWelland

e-Health

254

2.2 Personalisation, Contextualisation

and Error Trapping

The ASSISTANT product will be capable of

achieving a high degree of personalisation,

according to the user’s ability levels, and will be

able to be contextualised according to the user’s

situation. This contextualisation will take the form

of only providing information that is relevant to the

user, at the right time and in the appropriate format.

This will be achieved by filtering available data,

using information about the user’s preferences that

the user can set, and then communicating these data

to the user via audio, visual and haptic cues. The app

will achieve this by taking account of the user’s

context at any point in time, which will be defined

by their location, the stage they are at in the journey,

their mode of transport and the user’s stated personal

requirements and preferences.

The product will provide further personalisation

through its error trapping and remediation

functionality. The definition of an “error”, in this

context, is an unplanned situation caused by the user

not getting off the bus at the correct stop, for

example, or getting lost or anxious when searching

for a bus stop. “Error trapping” describes, for

instance, the process of detecting that the user is not

following the route intended. ASSISTANT’s

“design for failure” approach relates to both the

possibility of human error, and the failure of system

components, and so acknowledges the fragility of

mobile devices, the brittleness of digital information

and the limitations of location information provided

by Global Positioning System (GPS) technology.

The app will use location, system state, and user

interaction behaviour as a basis for Artificial

Intelligence based error detection. Once an error

type has been detected, the user model will

determine the appropriate mediation strategy. Such

mediation will be based on user-specific replanning

of routes, in the event of divergence from the

planned route, using pre-set personalised strategies.

3 HUMAN FACTORS

CONSIDERATIONS

A major effort has been made, during the project, to

ensure that the user interfaces and functionality of

the ASSISTANT system are designed to be as

suitable as possible for its target user group - older

people. An important component of ensuring the

accessibility of the product has been the filtering of

data to ensure that the user is provided with only the

information that is necessary, given the context in

which the information is provided, and the user’s

stated personal preferences. Care has been taken to

ensure the provision of an uncluttered screen, larger

buttons and an intuitive representation of screen

elements, and the design has also included a

reduction in the number of menus and drop-down

elements that are usually associated with

smartphones. As an example of the ASSISTANT

app’s user interface, Figure 2 shows two smartphone

screenshots, taken during the Prototype Phase.

Figure 2: Illustration of alternative means of providing

navigational information – compass-based (left) and map-

based (right).

The maximising of the accessibility of the product

has been made possible through the involvement in

the project of volunteers representing the device’s

target population group, at each stage of the

product’s development, i.e. the Concept Phase, the

Pilot Phase and, looking ahead to 2015, the

Prototype Phase. This has taken the form of

successive evaluation trials carried out in three

cities: Vienna, San Sebastián and Helsinki.

3.1 Evaluation Trials with End Users

A sample of 30 people, over the three cities, was

structured so that at least 90% of the participants

were public transport users – with half of the sample

being frequent users – and at least 50% of the

sample used a mobile ‘phone. All participants were

aged 55 or over, reflecting the envisaged market for

the ASSISTANT app, and there was an even gender

split.

The focus of the Concept Phase evaluations was

the design of the ASSISTANT user interface, with

there being the opportunity for further adjustments

TheAssistantProject-CreatingaSmartphoneApptoAssistOlderPeoplewhenTravellingbyPublicTransport

255

to be made after each subsequent feedback loop, up

to the use of prototypes in the third, and final, year

of the project, (2014-15). Concept Phase evaluations

were carried out with the aid of “life-size” mock-ups

of user interfaces presented on a computer screen.

These mock-ups enabled participants to interact with

each interface in much the same way that they would

interact with the touchscreen of a smartphone, with

the clicking of the various “buttons” provided taking

them to a different screen. Using this approach, it

was possible for the participants to, for example,

plan hypothetical journeys by public transport.

During the Pilot Phase, evaluations of a pilot

version of ASSISTANT were carried out, using

smartphones, in the three cities, in June 2014. The

main objective of this second wave of evaluations

was to gather feedback, from representatives of the

target user group, on the usability and the usefulness

of the system.

Each participant in the evaluation was asked to

create a route for a journey by public transport,

using the ASSISTANT route planner. The

participant was then asked to make the planned

journey, which entailed walking “the first kilometre”

to the bus or tram stop, then travelling using the

selected means of public transport, and then walking

“the last kilometre” to the selected destination.

During this task, a test facilitator “shadowed” the

participant, ensuring that the ASSISTANT system

functioned correctly, and being available to help in

case of problems. At the end of each evaluation

session, feedback on the participant’s experience

with the ASSISTANT system was obtained by

means of a semi-structured interview and an

evaluation questionnaire.

In terms of the type of information that the user

will require, it has been acknowledged that older

people require the same information when travelling

as anybody else. However, in addition, the user will

be provided with information on potential physical

barriers to travel, on the accessibility of specific

platforms and stations, and on staffing levels and the

availability accessible toilets etc., as far as data

availability allows.

Prior to these evaluation trials, a qualitative

questionnaire elicited demographic information

about the participants, and details of their use of

both public transport and current assistive

technologies, particularly mobile ‘phones and

smartphones. An important goal of this early phase

of the research was to gain insights into the use of

mobile communications technologies by older

people in their daily, travel-related routines. It was

important to establish the priorities, needs and

acceptance factors of this diverse group of people, so

that the design process could be guided from a user’s

perspective, using a needs framework.

3.2 Findings from Qualitative Research

and User Evaluations, to Date

A strong theme to emerge from evaluations

involving potential users has been people’s concern

for personal security when carrying and using an

expensive item of equipment such as a smartphone.

Particular concern has been noted at the prospect of

conspicuously using such a device on some urban

public transport networks, especially at night. This

represents encouraging feedback, in as much as a

major selling point of the ASSISTANT system will

be the facility for it to be used, with its options for

tactile and/or audible output, whilst safely concealed

in a pocket or a hand-bag.

Issues were also observed, among the research

participants, with performing tasks at the

intersection of two computer systems. Specifically,

this occurred when participants were asked to plan a

route on their PC, and then execute it using a

smartphone. ASSISTANT will address this potential

barrier to use of the project’s app by making sure

that both the PC and the smartphone are easy to use,

with seamless data transfer between the two and

transparent user interfaces.

A general conclusion that has been drawn from

the evaluation of alternative user interface designs is

that there is a marked preference for a simple,

uncluttered screen. For the app’s personal navigation

function, the majority of participants have expressed

a preference for step by step, text instructions,

although some have appreciated the facility to

switch to a map display. Where a map has been

used, requests have been made for a “You Are Here”

symbol indicating the user’s position. There has also

been a general preference shown for two levels of

zoom, whereby an overview map is followed by

more detailed instructions.

Negative feedback received from the evaluations

has included complaints about having to switch to

using reading glasses for accessing on-screen

information, and having to hold a smartphone device

in the hand whilst travelling – but this issue has

since been addressed by the use of Bluetooth

earpieces. Furthermore, the rate of providing spoken

information to the user has been slowed down to one

of 80% of “standard” speech, in order to make it

more accessible to older people.

Some participants have complained that the

touchscreen is too sensitive to light touches, which

ICT4AgeingWell2015-InternationalConferenceonInformationandCommunicationTechnologiesforAgeingWelland

e-Health

256

might be involuntary, and so this is an additional

design factor that will be taken into account during

the creation of the prototype.

A marked preference has been noted for the

actual name of a stop to be specified, whenever it is

identified as the stop where the user should get off

the vehicle. When the app indicates that the user

should board a vehicle, it has been noted by some

participants that the app should always indicate to

the user the name and number of the vehicle, and its

direction of travel or ultimate destination, (and

appropriate changes have been made to the system,

in time for the Prototype Phase evaluations taking

place in May 2015). There has been evidence to

suggest that some people are not comfortable with

instructions being given in terms of metres to a

required vehicle or destination.

Issues raised with reference to the route editing

function of the system, which will be designed to be

executed using a home PC, have included the

legibility of text for street names on map displays.

This has highlighted the importance of text

remaining of sufficient size to be legible, once the

user has changed the scale of the display, and this is

an issue that has been taken on board in the design

of the prototype system. A preference has also been

noted for using an address expressed as text as a

means of specifying a desired destination, as

opposed to selecting a point on a map display. The

least preferred mode of input has been demonstrated

to be the use of coordinates.

Participants found some of the terminology used

in the initial mock-up of the route editor to be

confusing – words such as “font” – so care has been

taken to use plain language in the design of

subsequent versions of the app. Generally, there has

been an appreciation for the facility to adjust

parameters such as font size, colour and contrast, as

well as output such as volume and haptic feedback.

4 THE MAIN CHALLENGES

RELATED TO THE SUCCESS

OF ASSISTANT

The main technical challenges for developing the

app have been concerned with the very basic

functions of ASSISTANT’s navigation feature,

particularly pin-pointing the user’s precise location

at any time and providing accurate routing

instructions in real time. In the context of the

pedestrian environment, positioning is particularly

difficult, given the challenges of “urban canyons”

and the fact that the device does not have the

structure of a road or rail network as a guide to

where the user might be. However, ASSISTANT

addresses the issue of intermittent reliability of

positioning information through its use of three

complementary means of guidance – map, compass

and distance tether. The app will also be constrained

according to the functional capabilities of chips and

other items of equipment that are commercially

viable for use in a smartphone.

An important finding from user evaluations to

date has been volunteers’ requests for being able to

create and change routes “on the hoof”, using their

smartphone, which, they felt, would constitute a

useful improvement to the app.

The provision of accurate and up to date

information on public transport services is another

prerequisite for the app’s viability, and this will, in

turn, depend upon the extent to which the local

transport operator is prepared to provide such

information for commercial use. How many city

authorities are prepared to embrace the “open data”

concept will determine how widely the ASSISTANT

app can be rolled out, which will be important to its

commercial viability.

5 SUMMARY

Evaluations of the ASSISTANT app have been

carried out using both Concept Phase mock-ups on a

computer screen, and a fully functional pilot version

of the system using smartphones. These evaluations

have featured samples of volunteers aged 65 and

older in Finland, Austria and Spain, with the

objective of assessing the usability and the

usefulness of the system for the target end user

population, but also with the aim of detecting system

bugs and errors. The data collected during these

evaluations have been crucial in helping

ASSISTANT’s technical partners to specify and

fine-tune the system to ensure that the final product

will meet the target population’s expectations and

needs.

Generally, the feedback gathered during the

Concept and Pilot Phases has been positive, and

participants have indicated that ASSISTANT would

be very helpful in enabling them to more easily and

more comfortably use public transport. Evaluations

using a pilot version of the app have indicated that

the ASSISTANT system can provide a helpful

service for older people when using public transport.

The general reaction has been that the system is

relatively easy to use, in the context of both the

TheAssistantProject-CreatingaSmartphoneApptoAssistOlderPeoplewhenTravellingbyPublicTransport

257

personal for initially planning the route, and the

smartphone with the ASSISTANT app during the

journey.

ICT4AgeingWell2015-InternationalConferenceonInformationandCommunicationTechnologiesforAgeingWelland

e-Health

258