Dementia Care

Vol 5 No 3 June/July 2016

Australian Journal of Dementia Care

design scenario set in the year 2030 as the

framework for my research. This scenario

focuses on a routine trip into the public

realm, undertaken with the aid of the

PRO-d by an imagined protagonist

named Gladys, a 72-year-old North

Melbourne resident with prodromal

dementia.

Gladys and the PRO-d

In this scenario, Gladys lives in a large

Victorian terrace just a short walk from

Errol Street. Three years ago she noticed

she had some difficulty in recalling

appointments, and often the names of

shopkeepers in her local area. On one

occasion she became lost while walking

home from the train station; a frightening

experience that made her reluctant to

leave the house for fear of getting lost

again. Gladys described her symptoms to

her GP. The doctor, recognising Gladys’

strong preference for living

independently, introduced her to the

PRO-d.

New design principles

The set of design principles developed

for this project underpin the PRO-d

proposal and have been created to

specifically addresses a gap in existing

design guides – the experience of

dementia in the public realm. These

Seven Design Principles for Dementia in

the Public Realm aim to provide a

practical basis for public realm design.

While by no means an exhaustive list, my

hope is that the principles may prompt a

broader conversation around inclusive

architecture and urban design.

I developed the principles by analysing

a public realm case study (Errol Street,

North Melbourne) and selectively

reinterpreting three recognised sets of

design principles: the 7 Principles of

Universal Design (Steinfeld 2012); 10

Principles of Designing for Dementia

(Fleming 2011) and PLANET (Person,

Location, Architecture, Nature, Energy,

Technology) (Chalfont 2007).

My Seven Design Principles for

Dementia in the Public Realm are:

1. Avoid streetscape symmetry and

uniformity. Avoid symmetrical

intersections and create a legible street

hierarchy by varying the width and

providing distinguishing features

(landmarks and/or latent

environmental cues) that are visible

from the street intersection.

2. Provide contrasting colour for changes

in level or surface treatment and to

delineate pedestrian vs vehicular

zones. Avoid high contrast ground

patterning.

3. At regular intervals, provide

universally legible signage (at a height

that is visually accessible for the

elderly, with appropriate colour

contrast and font) combining explicit

language and symbols.

This is

particularly important in areas that require

critical decisions (intersections) or have

potential sources of distraction (tram bells

or sirens).

4. Provide adequate crossing time at

intersections.

5. Provide frequent rest opportunities

that balance the provision of sunlight

and shade.

6. Provide public spaces (including green

spaces) that are small in scale and have

clarity of purpose.

7. Provide ‘edges’ (locations with both a

sense of protection or security, and a

view to surroundings – for instance, a

shaded public bench close by a tennis

court, or a window seat in a public

library) to offer an opportunity for

passive connection to nature and/or

involvement in community activity.

While any ambition to improve the

physical built environment is admirable,

cost and regulatory inertia are major

inhibitors to any major RL public realm

renovation. Anticipating this

conundrum, my research focused on

translating the Seven Design Principles

for Dementia in the Public Realm into

AR; hence, the creation of a fictional

augmented reality device, the PRO-d.

While not all principles could be fully

digitised – notably, ‘edges’ could only be

emphasised, not conjured – overall, the

proposed AR translation provides a

reasonably effective workaround.

The technology

The hypothetical PRO-d relies on five

existing technologies: two established

(the internet/mobile cellular network,

and the global positioning system, or

GPS) and three at critical points of

development at the time of writing

(depth-sensing cameras, automated

pattern/text recognition and the head-

mounted display, or HMD).

While the technologies needed to

create this device are on the cusp of

commercialisation (at the time of

writing), the PRO-d scenario is set in

2030 to allow for the arrival of a cohort of

people in their 70s and 80s, who (though

not ‘digital natives’) are sufficiently

familiar with smart devices that they are

able to manage the operation of the PRO-

d even if experiencing cognitive

difficulties.

The PRO-d would be assembled in this way:

Apair of conventional-looking spectacles

is fitted with a very small computer

processor (a system-on-chip, or SOC

processor as used in Google Glass) and a

depth-sensing camera. The depth-

sensing (or time-of-flight) camera is

similar in appearance to a typical smart

phone camera. However, as well as

recording two-dimensional images, the

depth-sensing camera captures terrain

data, generating and continually

updating a 3D model of Gladys’ physical

surroundings in real-time (Qualcomm

2014).

Meanwhile, the 2D images are

analysed by a ‘reading robot’ – a pattern

and text detection system developed to

support navigation for blind or visually

impaired people (Bulacu

et al

2008). This

textual data is merged with the 3D

terrain model to create a comprehensive

‘map’ of Gladys’ physical environment.

This virtual ‘map’ is paired with GPS

location data, then analysed within the

SOC by an algorithm operating

according to the design principles

previously discussed. The algorithm

identifies hazards and/or opportunities

in the physical environment and

generates an appropriate visual or audial

response, which is relayed to the eye

and/or ear of the user via the HMD. All

this happens in a fraction of a second, but

to Gladys, the process is instantaneous;

augmented reality merges seamlessly

with real life.

The PRO-d is activated by natural

language voice commands, like ‘I’m going

to the bakery’. The device operates on the

basis of minimal interference with the

user’s navigational approach unless

assistance is requested or deemed

necessary; for instance if the user departs

significantly from regular routes or asks

for help (‘Where is the bakery from here?’).

The device relays information to the

user primarily by integrating graphic

elements into the existing physical

environment. These graphic elements

conform to a visual language developed

according to a combination of universal

conventions and dementia-specific

requirements. For instance, colours have

been selected based on two criteria;

intuitiveness and ease of visibility.

Declining colour discrimination is a

common trait amongst the elderly, in

which colours become less saturated,

particularly colours with shorter

wavelengths (blue) (Ishihara

et al

2001;

Wijk

et al

2002). Thus, the PRO-d display

favours bright colours predominately in

the red to yellow range of the spectrum.

As red and orange are generally

associated with alarm or anger, pink and

yellow (still highly visible) are used for

less urgent prompts. Colours universally