COMMUNITY MEMORIES FOR SUSTAINABLE URBAN LIVING

Ellie D’Hondt and Matthias Stevens

BrusSense Team, Department of Computer Science

Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium

Keywords:

Pollution, Citizen science, Sustainability, Participatory sensing, Geo-localisation, Tagging, Mobile phones.

Abstract:

We propose a particular approach by which computer science can aid on the technological side of sustainable

development, which at the same time contributes to raising people’s awareness of the issues at hand. In

particular we develop a so-called community memory for urban environmental measurement surveys, focusing

on noise, microclimate and atmospheric pollution. Our goal is to gather data both in a quantitative way, using

participatory mobile sensing, and in a qualitative way, by social tagging. In this way we provide a method for

environmental monitoring which is complementary to current modelling approaches.

1 INTRODUCTION

The past decade an overwhelming amount of evidence

has accumulated for the fact that the current organisa-

tion of our societies is unmanageable with respect to

the available natural resources. In order to guaran-

tee our future well-being we fundamentally have to

change our lifestyle so that it becomes sustainable,

i.e. so that it achieves an ecological balance by avoid-

ing depletion of natural resources. A lot can and must

be done from the technological and policy side, and as

such sustainability is now a key issue ﬁguring in many

a government directive, company slogan and research

project. On the other hand we are still a long way off

a global transition in everyday lifestyle, as it is only

when people become fully aware of their precarious

ecological situation that one can expect the needed be-

haviour change.

A core ecological issue standing in the way of sus-

tainability is that of environmental pollution, in par-

ticular in urban environments. Combined with sea-

sonal conditions this can lead to critical situations,

with carbon dioxide, ozone and ﬁne particles being

the main concerns. Noise pollution is also a ma-

jor problem in urban environments, affecting human-

behaviour, well-being, productivity and health. Of

course there is already environmental monitoring go-

ing on, demanded by ambitious European norms

such as the European Environmental Noise Direc-

tive 2002/49/EC or guideline 1999/30/EG for Partic-

ulate Matter concentrations. Cities and governments

are already aware of the need for policy or planning

changes to satisfy these norms. However, these efforts

are hampered by two problems: one is the inadequacy

of current pollution measurements, the other the lack

of awareness of the average citizen.

The main issue with current assessment tech-

niques for environmental pollution is that they lack

spatio-temporal data granularity. This is because they

are based on computer simulations which rely on a

very limited amount of actual measurement data only.

As an example, a medium-sized city such as Brus-

sels has only thirteen measuring points for air qual-

ity, which, considering that air quality can change

from one side of the street to the other, is totally in-

adequate. With propagation models – based on gen-

eral statistics on trafﬁc ﬂows and industrial activity

– these local measurements are extrapolated to wider

areas to create pollution maps. These maps only

show expected global patterns and are often updated

only infrequently.

As a result, it is very hard to

evaluate the situation in a given neighbourhood, let

alone account for local or accidental noise or atmo-

spheric pollution problems such as road works, street

protests, etc., which are a real burden for citizens

none the less. Because current environmental mon-

itoring techniques are demanding at the levels of ex-

pertise, human resources, computation power, as well

as the quality of measuring devices, they are expen-

sive and not scalable enough to improve easily on

spatio-temporal granularity. As a consequence, there

is no convenient way to assess the individual pollu-

tion exposure of citizens, and indeed few efforts have

For example, EU norms demand noise maps to be up-

dated every ﬁve years.

D’Hondt E. and Stevens M. (2010).

COMMUNITY MEMORIES FOR SUSTAINABLE URBAN LIVING.

In Proceedings of the Multi-Conference on Innovative Developments in ICT, pages 77-80

DOI: 10.5220/0003039600770080

 SciTePress

been made for people-centric monitoring of the envi-

ronment (Stapelfeldt and Jellyman, 2003).

A second important disadvantage of current data-

gathering methods is the lack of human involved-

ness. Numerous international reports have expressed

the importance of the participation of all citizens,

at all levels, to move towards sustainable devel-

opment (United Nations Environment Programme,

1992; European Parliament and Council, 2002). In-

deed, as is the case with many issues affecting ur-

ban life, pollution cannot be tackled by policy makers

alone, as it requires consideration of the behaviour of

all citizens. In reality, however, citizen participation

often occurs only at the decision making level, and as

such involves only a limited number of citizens. The

problem is one of awareness. Pollution surveys are

carried out by a select group of people using expen-

sive and difﬁcult-to-handle equipment. As such most

citizens have no access to tools to estimate the quality

of their personal environment and how it is affected by

their behaviour and that of their peers. Moreover, cur-

rent environmental measurements are totally decou-

pled from the average citizen’s experience of the situ-

ation. Indeed physical measurement data only tell one

side of the story, and need to be augmented with sub-

jective, qualitative data that trace out the environment

as experienced by individuals.

The bottleneck here is

that existing techniques based on survey forms or in-

terviews by social scientists, do not yield enough data

to be representative on a large-scale. Furthermore be-

cause this data is collected separately it is hard to align

it with physical measurement data. As a result valu-

able information such as knowledge of the source of

pollution or subjective experience thereof, is entirely

absent.

In this paper we propose a particular approach

by which computer science can aid on the techno-

logical side of sustainable development, which at the

same time contributes to raising people’s awareness

of the issues at hand. In particular we exploit the

democratisation of technology facilitated by the Inter-

net of Things to realise an increase in and democrati-

sation of environmental information. The Internet of

Things’ vision of connectivity of anyone at anytime

to anything is largely supported by the widespread

use of mobile phones and advances in sensor tech-

nology. Our aim is to put these technologies to use to

enable digitally improved citizen science – the idea of

having volunteers without scientiﬁc training perform

research-related tasks such as observation, measure-

ment or computation – in the context of environmen-

For example in the case of noise the same intensity

of sound can give a totally different subjective experience

ranging from pleasant to aggravating.

tal monitoring. As the technologies we rely on are in-

trinsically people-centric, we tackle both the problem

of limited spatio-temporal data granularity and that of

the lack of human involvedness that are typical of cur-

rent simulation-based techniques.

2 PROPOSED SOLUTION

The technological framework we propose for al-

lowing high-granularity person-centric environmental

monitoring is that of community memories. A com-

munity memory (Steels, 2008) is a common ICT re-

source that citizens use to monitor their environment,

to visualise gathered data and to strategise about al-

ternatives for keeping it sustainable. We gather data

both in a quantitative way, using participatory mo-

bile sensing, and in a qualitative way, by social tag-

ging. Participatory sensing (Burke et al., 2006) ap-

propriates everyday mobile devices such as cellular

phones to form interactive, participatory sensory net-

works that enable public and professional users to

gather, analyse and share local knowledge. Social tag-

ging (Steels, 2006; Steels and Tisseli, 2008) augments

this information with tags, open-ended keywords en-

tered through mobile phones ( or through web inter-

faces) as meta-data. The type of data we envision

consists of measurements of environmental parame-

ters such as noise or pollutant concentrations com-

bined with tags for geographical location and time.

Geographic localisation requires sophisticated tech-

nologies such as GPS, GSM-based positioning (Chen

et al., 2006; Gonzalez et al., 2008), or localisation

through WiFi access points (Rekimoto et al., 2007).

On top of this users may tag measurement data with

subjective experiences or the suspected source of pol-

lution. For example, a user might indicate that noise

at a particular spot on the way home from work is

likely to originate from a building site. While there is

no restriction in principle, users tend to converge on

the same tags when they can see tags of others and

their aggregation in tag clouds (Cattuto et al., 2007).

Individual participatory sensing data will obviously

not have the same quality and accuracy as that ob-

tained from high-end environmental monitoring sta-

tions. However, we expect that this issue can be alle-

viated by collecting massively more data and by us-

ing techniques to calibrate and correlate those data. A

schematic representation of the underlying ICT archi-

tecture is given in Figure 1, with mobile device-based

data gathering components on the left-hand side and

the server-based community memory implementation

on the right-hand side. Mobile application software

provides an interface for the user, also allowing tag-

INNOV 2010 - International Multi-Conference on Innovative Developments in ICT

ging, either automatically (for spatio-temporal data)

or manually. It also takes care of data transmission

to a central server, which hosts the actual community

memory. At the server side more software takes care

of aggregating gathered data and visualising it on ge-

ographical maps or as tag clouds, as well as providing

an interface for users to interact with one another so

as to build up their community.

The community memory we propose serves both

as an instrument for supporting actions towards pol-

icy makers as well as a platform for making the av-

erage citizen aware of his urban environmental con-

ditions. That community memories have real value

both for the community itself as well as for policy

makers is proved by their success in situations going

from semi-nomadic Mbendjele Pygmies struggling

for the preservation of the rainforests of Congo (Hop-

kin, 2007) to handicapped people striving for easy-

to-navigate streets in Barcelona (Steels and Tisselli,

2008). Moreover the community memory acts as a

repository of environmental data at a much higher

granularity than that of current measuring methods,

which moreover includes person-centric information,

entirely lacking in present-day surveys (Brussels In-

stituut voor Milieubeheer, 2002). Therefore commu-

nity memories resolve both the granularity as well as

the awareness issues highlighted in the above. We

note that to scale up the centralised setup proposed

here one can rely on conventional server duplication

and load-balancing techniques.

To experiment with and evaluate the approach de-

scribed here we are involved in NoiseTube initia-

tive (Maisonneuve et al., 2009). The goal of this on-

going research project is to develop a participatory so-

lution to the monitoring and mapping of urban noise

pollution. Concretely a software platform is provided

that enables citizens to measure their personal expo-

sure to noise by using GPS-equipped mobile phones

as noise sensors. The system allows participants to

share geo-localised measurement data though a com-

munity memory website in order to contribute to col-

lective monitoring initiatives. These initial exper-

iments demonstrate convincingly the need and the

willingness of citizens to be involved, but so far they

have been limited in time and space, mostly due to

time and ﬁnancial constraints.

The creation of community memories is now tech-

nically feasible largely due to the fact that networked

electronic components are becoming more and more

wide-spread, versatile and affordable. Nevertheless

they require many state-of-the art technologies and

even act as a driver for additional advances in several

areas of computer science and engineering, such as

participatory sensing. Besides the NoiseTube project

other preliminary experiments in active participatory

sensing of the environment have been conducted,

partly organised by governmental organisations and

partly by citizen organisations (Ellul, 2008; Mun

et al., 2009). The idea of using commodity devices

such as mobile phones for scientiﬁc purposes and en-

vironmental monitoring in particular is also gaining

traction outside academia, as illustrated by coverage

in science and technology press and mainstream me-

dia (Patel-Predd, 2009; The Economist, 2009).

Configuration

- user preferences

- phone type

- calibration

Automatic

capture

- time: system clock

- locate: GPS/GSM/WiFi

- sense: on/offboard

Manual

capture

- social tags

Data Actions

- aggregation

- visualisation

(on map & tag clouds)

- improvement

- mining

- linking

Extra Tools

- user interaction

- data selection

- environmental log

device server

Figure 1: Community memory architecture.

3 CONCLUSIONS & FUTURE

WORK

In this paper we present the current status of our work

on developing participatory approaches to environ-

mental monitoring. We describe the limitations of

current techniques, the advantages of the participa-

tory approach and the technical challenges it brings.

NoiseTube, our prototype system, represents a work-

ing solution for participative monitoring of urban

noise pollution but remains under active development.

First, in order to evaluate and further improve the cur-

rent system, experimental medium-size deployments

with volunteering citizens are planned for the near fu-

ture. Subsequent larger-size experiments are needed

to validate the overall participatory approach to en-

vironmental monitoring. These efforts require us to

tackle several open challenges, such as developing

a procedure for estimating the credibility of aggre-

gated data as well as improving the quality of the

data through spatio-temporal interpolation or by com-

plementing measurements with propagation methods.

Second, we intend to build upon technologies devel-

oped and lessons learned to evolve NoiseTube into a

more powerful and versatile community memory plat-

form for assessing not only noise but also atmospheric

pollution. In this context sensing is more difﬁcult as

COMMUNITY MEMORIES FOR SUSTAINABLE URBAN LIVING

it involves off-board equipment which is also typi-

cally less stable. Third, once data sets become large

enough one can consider more sophisticated analy-

ses through data mining techniques, or by linking our

data repository with databases dealing with topics as

diverse as trafﬁc ﬂows, building schedules or disease

occurrences. Finally, within the larger context of sus-

tainability we cannot but consider the energy trade-

off of the technological developments we propose. In

particular, we need to assess the amount the power

our overall solution consumes for different numbers

of contributing and consulting users.

ACKNOWLEDGEMENTS

The ﬁrst author is supported by the Institute for the

encouragement of Scientiﬁc Research and Innovation

of Brussels (IRSIB/IWOIB) and the second by the

Flemish Fund for Scientiﬁc Research (FWO).

REFERENCES

Brussels Instituut voor Milieubeheer (2002). Plan voor

structurele verbetering van de luchtkwaliteit en de

strijd tegen de opwarming van het klimaat. (in Dutch).

Burke, J. A., , Estrin, D., Hansen, M., Parker, A., Ra-

manathan, N., Reddy, S., and Srivastava, M. B.

(2006). Participatory sensing. In World Sensor Web

Workshop (WSW

ı06) at ACM SenSys

ı06, October 31,

2006, Boulder, Colorado, USA.

Cattuto, C., Loreto, V., and Pietronero, L. (2007). Semi-

otic dynamics and collaborative tagging. PNAS,

104(5):1461–1464.

Chen, M. Y., Sohn, T., Chmelev, D., Haehnel, D., High-

tower, J., Hughes, J., LaMarca, A., Potter, F., Smith1,

I., and Varshavsky, A. (2006). Practical Metropolitan-

Scale Positioning for GSM Phones. In Dourish, P.

and Friday, A., editors, Proceedings of UbiComp 2006

(Orange County, CA, USA, September 17-21, 2006),

volume 4206/2006 of Lecture Notes in Computer Sci-

ence, pages 225–242. Springer.

Ellul, C. (2008). Creating Community Maps for the London

Thames Gateway. In RGS-IBG Annual International

Conference 2008.

European Parliament and Council (2002). Directive

2002/49/EC relating to the Assessment and Manage-

ment of Environmental Noise. Ofﬁcial Journal of the

European Communities, 18.7.2002:12–26.

Gonzalez, M. C., Hidalgo, C. A., and Barabasi, A.-L.

(2008). Understanding individual human mobility

patterns. Nature, 453(7196):779–782.

Hopkin, M. (2007). Conservation: Mark of respect. Nature,

448:402–403.

Maisonneuve, N., Stevens, M., Niessen, M. E., Hanappe,

P., and Steels, L. (2009). Citizen noise pollution mon-

itoring. In dg.o ’09: Proceedings of the 10th Annual

International Conference on Digital Government Re-

search, pages 96–103. Digital Government Society of

North America / ACM Press.

Mun, M., Reddy, S., Shilton, K., Yau, N., Burke, J., Es-

trin, D., Hansen, M., Howard, E., West, R., and Boda,

P. (2009). PEIR, the Personal Environmental Impact

Report, as a Platform for Participatory Sensing Sys-

tems Research. In MobiSys ’09: Proceedings of the

7th international conference on Mobile systems, ap-

plications, and services, pages 55–68. ACM.

Patel-Predd, P. (2009). Cellphones for Science. IEEE Spec-

trum, 46(2):16.

Rekimoto, J., Miyaki, T., and Ishizawa, T. (2007). LifeTag:

WiFi-Based Continuous Location Logging for Life

Pattern Analysis. In Hightower, J., Schiele, B., and

Strang, T., editors, Location- and Context-Awareness,

Third International Symposium, LoCA 2007, Oberp-

faffenhofen, Germany, September 20-21, 2007. Pro-

ceedings, number 4718 in Lecture Notes in Computer

Science, pages 35–49. Springer.

Stapelfeldt, H. and Jellyman, A. (2003). Using GIS in Noise

exposure analysis. In Proceedings of the 32nd Inter-

national Congress and Exposition on Noise Control

Engineering (Inter-Noise 2003; Seogwipo, Korea, Au-

gust 25-28, 2003).

Steels, L. (2006). Semiotic Dynamics for Embodied

Agents. IEEE Intelligent Systems, 21(3):32–38.

Steels, L. (2008). Community Memories for Sustainable

Societies. Technical report, Sony Computer Science

Lab - Paris. To be published on the occasion of the

20th Anniversary of Sony CSL, Tokyo spring 2008.

Steels, L. and Tisseli, E. (2008). Social Tagging in Com-

munity Memories. In Social Information Processing -

Papers from the 2008 AAAI Spring Symposium (March

26-28, 2008, Stanford University), number Techni-

cal Report SS-08-06, pages 98–103. The AAAI Press,

Menlo Park, California, USA.

Steels, L. and Tisselli, E. (2008). Interfaces for Commu-

nity Memories. In IUI 2008 Proceedings - 2008 In-

ternational Conference on Intelligent User Interfaces

(January 13-16, 2008, Canary Islands, Spain).

The Economist (2009). Sensors and sensitivity. The

Economist – Technology Quarterly, 391(8634):21–22.

Technology Quarterly issue.

United Nations Environment Programme (1992). Rio Dec-

laration on Environment and Development. Pro-

claimed at the UN Conference on Environment and

Development (Rio de Janeiro, Brasil, 3-14 June 1992).

INNOV 2010 - International Multi-Conference on Innovative Developments in ICT