IEEE Technology and Society Magazine - Winter 2013 - 52

the company, including employees,
customers, suppliers, universities, and
social networks, as in the Open Innovation theory described in [9].
Simply stated, Open Innovation
implies a paradigm in which companies open up their innovation
processes for the inflow and outflow of knowledge and information
to accelerate internal innovation
while exploiting external innovation [10]. From the point of view of
a private company, Open Innovation is not only about outsourcing
r&D but also about strategically
managing intellectual property
rights [11] by selling internally
unused knowledge to increase
revenues. Meanwhile, the firm
also acquires externally produced
knowledge that can improve the
speed, efficiency, and effectiveness
of its innovation process.
Since the early 2000s the term
"Smart Cities" has become increasingly more popular among city
council managers, media, and now
the public in general. Citizens are
the core of smart cities. They interact with the city, but new methodologies will be needed to facilitate
access to city sensors from the citizen point of view.

Deploying an Open Sensor
Data Platform
To achieve the goal of sharing and
open data to the public, some technical expertise on the part of citizens
will be required. A real environment -
or platform - will be needed to
achieve this goal. we will introduce
here some technical challenges and
considerations involved in building an Open Sensor Data platform.
Another important point is to guarantee the interconnection between
platforms, sometimes called federation, with other Open Data platforms. As a specific case, we detail
the technologies used to build the
Open Sensor Data platform developed within the Open Cities Project. After that, we explain how the
linked data concept can be applied
to Open Sensor Data.
52

Sensor Networks
Sensor networks act as feeders of
the Open Sensor Data platform.
wSNs are comprised of small,
cheap, power-efficient, wireless
sensor nodes that communicate
with each other to sense and monitor cities. Sensor nodes can perceive
various physical parameters of environmental conditions (e.g., temperature, light, vibration, pollution,
humidity, wind speed, etc.), which
can be used for a diverse and broad
set of applications, such as energy
management, healthcare, traffic
control, or to provide information
about environmental conditions in
cities. results obtained from wSNs
are based on a myriad of different
network technologies and solutions. Data provided by different
firms are generated in different formats, which makes it more difficult
for the various dedicated platforms
to interact in a way that provides a
smart decision (e.g., one platform
for traffic control, one for temperature monitoring, one for humidity
monitoring, etc., with none communicating to the others) [12] and [13].
Requirements
The Open Sensor Data platform
requirements are completely different from a system that only collects
static information. Since dynamic
data is real-time information, the
system that stores and publishes
this information must have specific
features to continuously succeed in
its task. The main requirements of
this system are the following:
a) Scalability: The platform must
offer the possibility of supporting an enormous amount of sensor networks without affecting
performance. This involves not
only the capacity required to
store a large amount of information but also the capacity to efficiently process that information.
b) reliability: The platform must
be robust, dynamically gathering sensor data and continuously providing it to the
applications that use it.

Low latency: The platform
must offer a short response
time to consumer applications, which will help them
to provide near real-time
performance.
d) Standardized formats: The
stored information must be
presented to consumer applications in an accepted format.
(we recommend following
directives from the w3C).
e) Standardized
connectivity:
The platform must offer a
standard API to allow sensor
networks and consumer applications to connect to it.
Before implementing an OSN
Platform we have to deal with how
to cope with the large number of
messages that are collected and
stored, and how to serve them in
real time to possibly thousands of
consumers.
To design our platform we
have analyzed the cases of Twitter
and Netflix, which have similar
requirements. For the first question
Twitter [14] can be a good case reference, as there are millions of people sending short messages. Twitter
can also be considered for the second question, as it is one platform
that serves real-time information
for each individual user based on
the sensors that the user has previously selected. The Netflix video
service platform is also a good
example [15]. Netflix is an example of a platform with low-latency
requirements, as it is designed for
video storage and reproduction.
Therefore, it may teach us how to
handle the exponential increase of
storage capacity as more information is deposited.
Both Twitter and Netflix use NOSQL databases [16] as a technical
solution for our two main questions.
NO-SQL tables follow a decentralized architecture composed of
multiple servers' clusters; data is
replicated in each cluster. This type
of database solves the scalability,
reliability, and low-latency problems. Scalability is solved because

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

wINTEr 2013

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - Winter 2013