Identifying Points of Interest for Elderly in Singapore through Mobile

Crowdsensing

Marakkalage Sumudu Hasala

, Billy Pik Lik Lau

, Viswanath Sanjana Kadaba

, Balasubramaniam

Thirunavukarasu

, Chau Yuen

, Belinda Yuen

and Richi Nayak

Engineering Product Development, Singapore University of Technology and Design, Singapore

Lee Kuan Yew Centre for Innovative Cities, Singapore University of Technology and Design, Singapore

School of Electrical Engineering and Computer Science, Queensland University of Technology, Australia

Keywords:

Big Data, Crowdsensing, POI, Voronoi Diagrams, Heat Maps.

Abstract:

This paper introduces a crowdsensing approach to identify the points of interest (POI) among the elderly

population in Singapore. We have developed a smartphone application, which passively collects sensors’

information (e.g. GPS location) on users’ mobile devices. Using such information, we can identify popular

regions and places among the elderly that could be useful for city planner in preparation for aging population.

Our results demonstrate different check-in patterns of various POI, and the elderly spend nearly 70% of non-

home duration around their neighborhood.

1 INTRODUCTION

Elderly population is a key component in any soci-

ety. In Singapore, the number of citizens aged 65 and

above is increasing rapidly. In fact, the resident Old-

Age Support Ratio (persons aged 20 − 64 years per

elderly aged 65 years and over) has signiﬁcantly re-

duced from 9 to 5.7 over the years of 2000 to 2015

respectively (Department of Statistics, 2016). Rising

elderly population is a challenge to address since their

requirements and interests should be understood in or-

der to provide them a better quality of life. According

to (Capella and Greco, 1987), elderly have tendency

of travelling around in their leisure time, which leads

to the thinking that it is important to identify their

preference of places to visit. In order to achieve such

information we need to reach the elderly and com-

municate to gather required information from them.

Rather than conducting verbal or written surveys, the

more convenient approach is to utilize the ubiquitous

nature of smartphones among the elderly as a resource

to collect data from heterogeneous sensors available

in smartphones. According to study in (Pang et al.,

2014), elderly in Singapore have positive attitude to-

wards smartphones and have found them to be use-

ful and entertaining. Hence, a smartphone applica-

tion based approach has the potential in quantitatively

identifying elderly lifestyle. The work presented in

this paper, adopts a smartphone based mobile applica-

tion as the tool to better understand the daily activity

of an elderly. Using the smartphone sensor informa-

tion collected through mobile application, we prove

that it is possible to identify the Regions of Interest

(ROI) and Points of Interest (POI) of elderly in Sin-

gapore.

The remaining sections of this paper are structured

as follows. Section 2, discusses the literature review

and related work done previously. Next, Section 3

discusses the methodology which is used to collect

and analyse data from elderly. Section 4 describes the

insights about ROI from the analysed data. Section 5

presents POI extracted from each ROI. Finally, Sec-

tion 6, concludes the paper with a summary of ﬁnd-

ings, and possible future work.

2 LITERATURE REVIEW

There has been previous research done to track user

locations using mobile applications. Tourist tracking

through smartphone location data (Viswanath et al.,

2014) has been done in order to identify travel sites

of tourists and to rate tourist sites based on user sur-

veys conducted through the mobile application. In

the context of smart homes, previous work has been

carried out for proactive environment change for el-

derly (Helal et al., 2003), cloud based real-time activ-

Hasala, M., Lau, B., Kadaba, V., Thirunavukarasu, B., Yuen, C., Yuen, B. and Nayak, R.

Identifying Points of Interest for Elderly in Singapore through Mobile Crowdsensing.

DOI: 10.5220/0006309300600066

In Proceedings of the 6th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2017), pages 60-66

ISBN: 978-989-758-241-7

ity tracking in home (Fahim et al., 2012) which helps

to remotely monitor elderly who stay at home (Yassin

et al., 2017), and RFID based location tracking (Kim

et al., 2013) (Liu et al., 2015). An adaptable interface

for smartphones which is customized to the elderly

has been developed in (Arab et al., 2013). Helping

people with limited mobility (e.g. elderly) by sug-

gesting accessible urban paths using crowdsourced

data has been done in (Mirri et al., 2014). A pas-

sive information gathering system using mobile ter-

minals such as smartphone applications, emails, and

phonecalls, has been done to gather knowledge and

experience from elderly and transfer to younger gen-

eration (Hiyama et al., 2013). In (Ghiani et al., 2013),

a platform has been developed to support elderly to

stay active in life in order to improve their healthy

living.

To the best of our knowledge, there is no previous

work done to identify the ROI and POI of elderly, us-

ing smartphone based mobile applications. Our focus

is not only for the elderly who stay at home, but also

for the elderly who prefer to travel around their neigh-

bourhood or further. Therefore, this paper presents an

approach to understand lifestyle of elderly, using the

data collected through smartphone applications, and

analysing them.

3 METHODOLOGY

A user-friendly smartphone application that runs on

Android platform, is designed and developed (SUT-

DDev, 2015) to collect data from elderly. More than

100 people, the majority aged 50 and above, who re-

side in Bukit Panjang area of Singapore participated

in the study by installing the application on their mo-

bile phones. In this paper, we reported only the data

from 50 participants, whose collected data is consis-

tent throughout the study period. There are 3 ma-

jor stages in the approach of identifying ROI and

POI, namely, Mobile Application Development, El-

derly Data Collection, and Big Data Analysis.

3.1 Mobile Application Development

The smartphone application, which is used to collect

data from participants is an improved version of the

application introduced in (Viswanath et al., 2014).

The application collects data such as GPS location

(including latitude, longitude, and accuracy), activity

data from Google API, noise level of the environment,

device battery level, and light sensor values. In the

process of identifying ROI and POI, in this paper, we

utilized only the location related data. In future work,

our intention is to utilize all the collected information

in order to extract deep insights about user behaviour.

3.2 Elderly Data Collection

The recruitment of target participant group for the

study (elderly who use Android smartphones and re-

side in Bukit Panjang area of Singapore) was done

in public locations around Bukit Panjang. All the

data were collected for one month of minimum pe-

riod with the written consent from every participant in

the study. Participants’ personal identity is used only

during registration. The data collection and analysis

is done through a unique auto-generated device ID to

ensure privacy of the participants. Figure 1 shows the

ﬂow of elderly data collection.

Figure 1: Elderly Data Collection.

3.3 Big Data Analysis

Aquired data from the mobile phones of the elderly

need to be analysed in order to gather meaningful in-

sights. Data analysis was done using back-end Java

application and MATLAB software package. Af-

ter the data is analysed, tools such as Google Maps,

Voronoi Diagrams, and Heat Maps are used to visual-

ize the gathered insights, so that it is possible to iden-

tify the ROI and POI.

Data pre-processing is done in 2 major stages such

as (1) Denoise which removes duplicate data records

and anomalies occured due to location accuracies, and

(2) Time sync which aligns data from different tables

in relevant time slots.

After pre-processing stage, the raw location data

were clustered using k-means clustering technique

(Lau et al., 2017) and generated a summary database

for each user, which contains the parameters as shown

in Table 1. A total 1781 number of clusters were gen-

erated from raw data of all 50 users.

Identifying Points of Interest for Elderly in Singapore through Mobile Crowdsensing

Table 1: Parameters in the summary database.

Parameter Remarks

ID Unique user ID

Gender Male/Female

Age User’s age

Location Latitude and Longitude of cluster location

Purpose Purpose of visit (e.g. H-Home, G-Grocery, W-Work)

Duration Time spent in particular cluster

Visits Number of visits to particular cluster

Total Time Duration of ﬁrst and last record from user

4 REGIONS OF INTEREST

ROI analysis is done in three categories. First, we

identify the percentage of home stay duration of each

user out of their total participation time. Next we

identify the ROI across user neighbourhood, which

is Bukit Panjang, Singapore. Further, an analysis is

done across Singapore to identify islandwide ROI.

ROI are analysed using a technique called Voronoi

Diagrams. In a Voronoi diagram, a plane is parti-

tioned into n points such that each polygon contains

one seed and every point in each polygon is closer to

the seed inside each polygon than to any other one.

These polygons are called Voronoi zones (Aurenham-

mer, 1991).

4.1 Home Stay Duration

Figure 2 shows the percentage of time each partici-

pant spent at home, out of his/her total participation

duration for data collection. It also shows the age and

gender of each of the participant. The oldest partici-

pant is 75 years old, and the average age for partici-

pants is 60 years. There are 28% of male, and 72% of

female participants. It can be seen that a large num-

ber of participants spend a large amount of their time

at home, in particular 44% of participants stayed at

their homes for more than 21.6 hours a day on av-

erage, 22% of participants stayed at their homes be-

tween 19.2 and 21.6 hours a day on average, 8% of

participants stayed at their homes between 16.8 and

19.2 hours a day on average, 10% of participants stay

at their homes between 14.4 and 16.8 hours a day

on average, and 16% of participants stayed at their

homes less than 14.4 hours a day on average. It can

be noticed that participant IDs 10 and 44, have sub-

stantially lower home stay duration, when compared

to others. One possible reason for such a result is they

switch off their mobile phones at night.

0 5 10 15 20 25 30 35 40 45 50

Users

100

120

Percentage

Age (Years)

Percentage Home Stay Duration and Ages of Users

Female

Male

Age

Figure 2: Percentage of home stay duration, gender, and age

of each user.

Figure 3: Voronoi zones and the number of participant

homes in each Voronoi zone across Bukit Panjang.

4.2 Across Neighbourhood

We divided Bukit Panjang area into 12 Voronoi

Zones, in order to get an understanding about partici-

pants’ home region since all of them are residents of

Bukit Panjang. The Voronoi zone distibution, zone

number, and number of participant homes in each

Voronoi zone is shown in Figure 3.

(a) Percentage of visits. (b) Percentage of durations.

Figure 4: Voronoi zones in Bukig Panjang area and percent-

ages of non-home zone visit and durations.

Figure 4a shows the percentage of visits in each

region and Figure 4b shows the percentage of dura-

tions in each region. In both Figures, the inner cir-

cles display the Voronoi zone number and percentage

of visits/duration accordingly. The outer rings show

the percentages of aforementioned percentages con-

tributed by non-home zone participants.

SMARTGREENS 2017 - 6th International Conference on Smart Cities and Green ICT Systems

From the Figures, it can be observed that visits and

durations in zone number 2 (Senja Cashew CC) and

zone number 7 (Junction 10) have 100% of non-home

zone participants, which means, all the participants

who visit and spend time in those two zones are resid-

ing outside them. Especially, none of the participant

homes are located in zone number 7. Zone number

2 also shows the highest percentage duration, while

zone number 3 (Masjid Al-Iman) shows the highest

percentage of visits.

Another interesting observation is that, zone num-

ber 3 has higher percentage of visits when compared

to percentage of duration of the same. The reason for

such a result is, because the Bangkit LRT station is

located inside zone number 3. In terms of percentage

duration, zones 2 and 6 are popular among non-home

zone participants. Zones 5 and 9 are popular among

home zone participants. The reason for such popular-

ity is because those zones include POI such as Bukit

Panjang CC (zone 6), Senja Cashew CC (zone 2), and

Fajar Market (zone 9).

4.3 Across Singapore

The map of Singapore is divided into 9 Voronoi zones

in order to understand what regions across Singapore

are in interest of users in terms of number of visits and

duration of time spent. The Voronoi seeds are selected

to cover the main regions of entire Singapore. The

Voronoi zone distibution and zone names are shown

in Figure 5.

Figure 5: Voronoi zones across Singapore.

Figure 6 shows the percentage number of visits

and durations for each region across Singapore. It

can be observed from the Figure that region number

3 (Bukit Panjang) is the most popular among partic-

ipants, in terms of both visits and duration. In fact,

53.5% of visits and 68.6% of durations belong to that

region. The reason for such popularity is because all

the participants are residing in Bukit Panjang area and

they tend to visit places nearby their homes. Apart

from that, regions 1, 4, 5, and 9 are popular regions.

However, regions 1 and 9 tend to have shorter dura-

tion than visits.

1 2 3 4 5 6 7 8 9

Regions

Percentage

Visits

Duration

Figure 6: Percentage of number of visits and duration across

regions in Singapore.

5 POINTS OF INTEREST

Next, we analyze in deeper details for each speciﬁc

POI, we have performed POI extraction and ranked it

according the number of participant and time stayed.

For privacy reason, we have removed participants’

home and work places. Later, we have clustered com-

mon POI among the participant using density based

clustering with radius of 100m (hence the POI indi-

cated later on include also nearby related POI). After

clustering process done, we have obtained 980 POIs

from the collected data.

5.1 POI Correlates with Number of

Participants and Stay Time

Table 2: POI Ranking based on Number of participants

Checked In.

No. Point of Interest

Percentage of

Participants

visit

Average Hours

Spent Per

Participant

Per Visit

1 Bukit Panjang Hawker Center 68% 3.7

2 Bukit Panjang Plaza 56% 2.9

3 Bukit Panjang Community Center 34% 1.9

4 Fajar Shopping Centre 30% 2.5

5 Senja Cashew Community Club 28% 4.4

6 Holland Bukit Panjang Town Council 28% 2.2

7 Block 125 Neighborhood Stores 26% 3.0

8 Lot One Shopping Mall 24% 1.7

National Healthcare

Group Pharmacy - Choa Chu Kang

22% 1.7

10 Chinese Temple at Bencoolen Link 22% 0.9

11 Greenridge Shopping Centre 20% 2.5

12 Ten Mile Junction Mall 20% 2.2

13 Plaza Singapura 18% 2.6

14 Causeway Point 18% 1.6

15 Block 172 Neighborhood Stores 16% 2.4

16 Zheng hua Community Club 16% 2.0

Teck Whye Ave

Neighborhood Stores

16% 1.3

19 Temple Street at China Town 16% 1.2

20 Phoenix Road Shop Lots 14% 2.6

Orange - Within Bukit Panjang Area, Blue - Within 5KM range of Bukit Panjang, Purple

- 5KM away from Bukit Panjang

We have plotted the heat maps for the participants’

based on the number of check in across Singapore as

displayed in Figure 7. The heat zones are shown in red

color, where the yellow color indicates the places, that

less participants visited. High heat area are circled

and pointed out int the maps, and we do not include

Identifying Points of Interest for Elderly in Singapore through Mobile Crowdsensing

Histogram on Check In Hours Total Time Spent on Day of Week

(a-1)Bukit Panjang Community Center

1AM 6AM 12PM 6PM 12AM

Time, from 0000 to 2359 Hours

Number of Check Ins

(b-1)Senja Cashew Community Center

1AM 6AM 12PM 6PM 12AM

Time, from 0000 to 2359 Hours

Number of Check Ins

(a-2)Bukit Panjang Community Center

SUN MON TUE WED THU FRI SAT

Day of Week

Total Time Spent (Hours)

(b-2)Senja Cashew Community Center

SUN MON TUE WED THU FRI SAT

Day of Week

Total Time Spent (Hours)

(c-1)Bukit Panjang Plaza

1AM 6AM 12PM 6PM 12AM

Time, from 0000 to 2359 Hours

100

Number of Check Ins

(d-1)Fajar Shopping Center

1AM 6AM 12PM 6PM 12AM

Time, from 0000 to 2359 Hours

100

Number of Check Ins

(c-2)Bukit Panjang Plaza

SUN MON TUE WED THU FRI SAT

Day of Week

Total Time Spent (Hours)

(d-2)Fajar Shopping Center

SUN MON TUE WED THU FRI SAT

Day of week

Total Time Spent (Hours)

(e-1)Bukit Panjang Hawker Center and Market

1AM 6AM 12PM 6PM 12AM

Hours, from 0000 to 2359 Hours

100

150

200

250

Number of Check Ins

(e-2)Bukit Panjang Hawker Center and Market

SUN MON TUE WED THU FRI SAT

Day of Week

100

Total Time Spent (Hours)

Note:

Fig. 8(a-1), (a-2), (b-1), and (b-2) - Similar Pattern, the difference is time spent on Sunday and Monday.

Fig. 8(c-1) is Gaussian shape histogram, while Fig. 8(c-2) is two peaks histogram; both have one or two ﬁxed participants, who always visit those POI.

Fig. 8(e-1) and (e-2) - morning place, and Monday no market so less, other days balanced

Figure 8: Check In Time and Time Spent for the following POI (Top to Bottom, Left to Right): (a) Bukit Panjang Community

Center, (b) Senja Cashew Community Center, (c) Bukit Panjang Plaza, (d) Fajar Shopping Center, and (e) Bukit Panjang

Hawker Center and Market.

Bukit Panjang Area as it is nearby to the home of par-

ticipants. We have highlighted a few high heat zone

area using circle and labeled it accordingly. We pre-

sented in Table 2 the top POI and time spent per visit

for the elderly, which the ranking is evaluated based

on the number of participants checked in. The POI

highlighted in orange indicates the POIs are located

within the Bukit Panjang area, where the blue color

indicates POI are within the 5 kilometers of the Bukit

Panjang area (nearby POI). Lastly, the purple color

indicates the POI, that are located outside of Bukit

Panjang area.

Based on observation, we noticed the top POI con-

sists of locations from Bukit Panjang area. Hawker

center and community center are the most common

China Town

Bugis

Lot One

Shopping Mall /

Chua Chu Kang MRT

Teck Whye Ave /

Chua Chu Kang

Community Club

Marina Square

City Plaza /

One KM Mall

Buono Vista

MRT Station

Jurong East,

J-Walk Area

Marsiling MRT

Shopping District

Woodlands MRT

Jurong Point

Clementi

MRT Station

Ikea

Alexandra

Ochard

Road

Limbang Park /

Choa Chu Kang

Sports Hall

Sembawang

God of Wealth

Temple

Figure 7: POI heat maps across Singapore.

places among the participants, where at least 1 hour

55 minutes per visit time spent is recorded. Then

the ranking followed by the shopping mall and neigh-

bourhood stores. Note that, a high density block

(HDB) can be a multi purposes places such as Rank

7, Block 125, which have grocery store and traditional

medical center at ﬁrst storey of the building. So, it is

not a uncommon situation in Singapore’s highly den-

sity new town area development. Participants have

spent an average of 2.5 hours in the Bukit Panjang

shopping mall area (Bukit Panjang Plaza, Fajar Shop-

ping Centre, Greenridge Shopping Centre, and Ten

Milt Junction Mall) per visit.

Next, we discuss the POI, located outside Bukit

Panjang area. As shown in Figure 7, there are a

handful of location outside of Bukit Panjang area dis-

tributed across central area (China Town and Bugis),

west south area(Jurong and Clementi), Bukit Panjang

nearby area, and north area (Woodlands). A top com-

mon area among the elderly is National Healthcare

Group Pharmacy, where it is located 2 kilometers

away from Bukit Panjang area. Other POI area are

mostly shopping malls and temples, which can be re-

lated to local cultures and religion.

In a nutshell, participants have the tendency to

visit nearby amenities for daily convenience. It can be

highly related to the transportation accessibility and

SMARTGREENS 2017 - 6th International Conference on Smart Cities and Green ICT Systems

travel time. However, when it comes to shopping, re-

ligious, and medical purposes, participants are more

likely to travel further.

5.2 POI Correlates with Check In Time

and Time Spent

After obtaining the POI ranking, we investigate the

check in time and time spent for each POI and the

results are displayed in Figure 8. We have divided the

analyzed data into two categories, which are check

in time according to hour of the day and time spent

according to the day of the week.

First, we analyzed the top POI from previous sec-

tion, which are Bukit Panjang Hawker Center and

Market. Based on the observation, we can consider

it as a morning place, where morning (8am - 10am)

have a lot check in compared to other hours. By com-

bining the day spent for different time of the week

reveals that Monday have less time spent there com-

pared to other days. The main reason is the operating

commercial hour of local market, where Monday is

the rest day. By compiling the information collected,

it implies Bukit Panjang Hawker Center is a place,

where most elderly would go to either secure their

grocery or other personal purposes. Note that, partici-

pants could have go the Bukit Panjang Hawker Center

for different purposes, and we only consider majority

purpose collected from the survey.

Next, we discuss about the shopping complex,

that majority elderly have checked in during the data

collection subsequently after Bukit Panjang Hawker

Center. There are two shopping complexs studied

here, which are Bukit Panjang Plaza and Fajar Shop-

ping Center. The pattern for both shopping complex

is diverse, that for Bukit Panjang Plaza is in Gaussian

distribution shape and Fajar Shopping Center is in two

peaks shape. It shows checked in time might highly

correlate with the location of the shopping mall. For

example, Bukit Panjang Plaza is Mass Rapid Transit

interchange, which is presumably have more checked

in compared to Fajar Shopping Center. We noticed

one or two participants that frequently spend their

time there and contribute a lot of hours.

Another common POI for the participants would

be community center. It is observed that both Bukit

Panjang and Senja Cashew Community center have

similar check in patterns, where morning and evening

are the high checked in time for the elderly. How-

ever, by comparing the number of hours spent for in

a week, there is a signiﬁcant variance between Bukit

Panjang Community center and Senja Cashew Com-

munity club. Bukit Panjang Community center has

more check in in Monday and Sunday and we believe

the reason could be linked with the activity a commu-

nity center offered and transportation accessibility to

reach there.

In conclusion, we noticed participants are likely to

go to Bukit Panjang Hawker Center and Community

Center in the morning. Such pattern can be differ-

ent subject to commercial operating hours and activ-

ities offered. However, contrast to community center

and hawker center, the visit pattern of participants on

shopping mall is highly correlates with the location of

the mall itself.

5.3 POI Correlates with Gender and

Working Type

The heatmaps according to different gender and

working type are presented in Figure 9. The highly

active gender corresponding to the working type are

highlighted with red bold boxes.

Working - female

Working - male

Non working - female

Non working - male

Figure 9: working type w.r.t. gender (a) Non working fe-

male, (b) Non Working male, (c) Working female, and (d)

Working male. (Top to bottom and Left to right).

First, we observed non working female are more

active than non working male elderly. A lot non work-

ing female elderly went to the center area for shop-

ping purposes. To name a few common spot for the

female participants are Bugis, Marina Square, China

Town, Orchard Road, and etc. Based on previous

data, we can draw a conclusion that majority of the

POI in central area are highly consists of shopping

malls etc. Contrast to female participants, non work-

ing male participants travel less and mostly spend

their time in Bukit Panjang nearby POI. However, we

noticed participants are willing travel more than 10

kilometers for religious purposes.

In contrast, working male participants are more

active that the female participants in term of traveling

around Singapore when they are working. It might be

correlates to the work nature and we yet to investigate

such occurrence. As for female participants, the heat

of maps is shift towards central and bottom part of

Identifying Points of Interest for Elderly in Singapore through Mobile Crowdsensing

the Singapore, where the male participants are spread

evenly.

Lastly, non working female and working male are

most active categories compared to working female

and non working male participants. It is hard to de-

termine what is the main factor that segregate the be-

havior of male and female according to their working

type. One things can be studied in our future works

maybe can correlates their age and their hobby to de-

termine the actual factor of separation.

6 CONCLUSIONS

In conclusion, this paper introduces an approach to

identify ROI and POI where elderly in Singapore pre-

fer to visit and spend their time. Using smartphone

based mobile application, it is possible to collect de-

tailed information with quantitative measurement.

In future, our intention is to identify the trans-

portation mode (e.g. walking, taking bus, taking train)

of the users, and proﬁle each user by creating a user

mobility proﬁle. The multi-sensor data collected from

smartphones make it possible to implement sensor fu-

sion techniques to determine high level information

about user behavior.

ACKNOWLEDGMENTS

Authors would like to express their sincere gratitude

to all the participants in the study who allowed us to

collect their location information. We would like to

thank everyone who spared their time to help us in

the process of elderly recruitment. This research is

supported by the Lee Kuan Yew Centre for Innova-

tive Cities under Lee Li Ming Programme in Aging

Urbanism.

REFERENCES

Arab, F., Malik, Y., and Abdulrazak, B. (2013). Evalu-

ation of phonage: an adapted smartphone interface

for elderly people. In IFIP Conference on Human-

Computer Interaction, pages 547–554. Springer.

Aurenhammer, F. (1991). Voronoi diagrams a survey of a

fundamental geometric data structure. ACM Comput-

ing Surveys (CSUR), 23(3):345–405.

Capella, L. M. and Greco, A. J. (1987). Information sources

of elderly for vacation decisions. Annals of Tourism

Research, 14(1):148–151.

Department of Statistics, S. (2016). In Statistics Singapore

- Resident Old-Age Support Ratio. Sing Stat.

Fahim, M., Fatima, I., Lee, S., and Lee, Y.-K. (2012). Daily

life activity tracking application for smart homes us-

ing android smartphone. In Advanced Communication

Technology (ICACT), 2012 14th International Confer-

ence on, pages 241–245. IEEE.

Ghiani, G., Manca, M., Patern

o, F., and Santoro, C. (2013).

Towards an architecture supporting social, adaptive

and persuasive services for active elderly. In CASFE,

pages 36–41.

Helal, S., Winkler, B., Lee, C., Kaddoura, Y., Ran, L.,

Giraldo, C., Kuchibhotla, S., and Mann, W. (2003).

Enabling location-aware pervasive computing appli-

cations for the elderly. In Pervasive Computing and

Communications, 2003.(PerCom 2003). Proceedings

of the First IEEE International Conference on, pages

531–536. IEEE.

Hiyama, A., Nagai, Y., Hirose, M., Kobayashi, M., and

Takagi, H. (2013). Question ﬁrst: Passive interaction

model for gathering experience and knowledge from

the elderly. In Pervasive Computing and Communica-

tions Workshops (PERCOM Workshops), 2013 IEEE

International Conference on, pages 151–156. IEEE.

Kim, S.-C., Jeong, Y.-S., and Park, S.-O. (2013). Rﬁd-based

indoor location tracking to ensure the safety of the el-

derly in smart home environments. Personal and ubiq-

uitous computing, 17(8):1699–1707.

Lau, B. P. L., Hasala, M. S., Viswanath, S. K.,

Thirunavukarasu, B., Yuen, C., Yuen, B., and Nayak,

R. (2017). Extracting point of interest and classifying

environment for low sampling crowd sensing smart-

phone sensor data. CoRR, abs/1701.03379.

Liu, R., Huski

c, G., and Zell, A. (2015). On tracking dy-

namic objects with long range passive uhf rﬁd using a

mobile robot. Int. J. Distrib. Sen. Netw., 2015.

Mirri, S., Prandi, C., Salomoni, P., Callegati, F., and Campi,

A. (2014). On combining crowdsourcing, sensing

and open data for an accessible smart city. In 2014

Eighth International Conference on Next Generation

Mobile Apps, Services and Technologies, pages 294–

299. IEEE.

Pang, N., Zhang, X., Vu, S., and Foo, S. (2014). Smart-

phone use by older adults in singapore. Gerontech-

nology, 13(2):270.

SUTDDev (2015). In City - Location tracking application.

Google Play.

Viswanath, S. K., Yuen, C., Ku, X., and Liu, X. (2014).

Smart tourist-passive mobility tracking through mo-

bile application. In International Internet of Things

Summit, pages 183–191. Springer.

Yassin, A., Nasser, Y., Awad, M., Al-Dubai, A., Liu, R.,

Yuen, C., Raulefs, R., and Aboutanios, E. (2017). Re-

cent advances in indoor localization: A survey on the-

oretical approaches and applications. IEEE Commu-

nications Surveys Tutorials, PP(99):1–1.

SMARTGREENS 2017 - 6th International Conference on Smart Cities and Green ICT Systems