Over the past decade, ‘smart’ cities have capitalized on new technologies and insights to transform their systems, operations and services. The rationale behind the use of these technologies is that an evidence-based, analytical approach to decision-making will lead to more robust and sustainable outcomes. However, harvesting high-quality data from the dense network of sensors embedded in the urban infrastructure, and combining this data with social network data, poses many challenges. In this paper, we investigate the use of an intelligent middleware – Device Nimbus – to support data capture and analysis techniques to inform urban planning and design. We report results from a ‘Living Campus’ experiment at the University of Melbourne, Australia focused on a public learning space case study. Local perspectives, collected via crowdsourcing, are combined with distributed and heterogeneous environmental sensor data. Our analysis shows that Device Nimbus’ data integration and intelligent modules provide high-quality support for decision-making and planning.
Digital Advertising Lecture for Advanced Digital & Social Media Strategy at U...
Distributed and heterogeneous data analysis for smart urban planning
1. Distributed and heterogeneous
data analysis for smart urban
planning
Eduardo Oliveira
Michael Kirley
Tom Kvan
Justyna Karakiewicz
Carlos Vaz
2. Outline
• Living Campus Project
• Research Ques:ons
• Related Work: an introduc:on to middleware
• Device Nimbus
• Case Study: proof of concept demonstra:on
• Conclusions and Future work
3. Living Campus
University campuses represent an urban space
that in many circumstances reflects what is
happening on a larger scale across a city.
5. Living Campus: an interdisciplinary perspec:ve
[architecture]
• Architects, planners, and urban designers typically require access to spa:al
and temporal data, which considers how people perceive, behave and
interact with their environment
• Data collec:on and analysis is rarely pitched at the `micro’ scale
[computer science]
• PaSS = People as Sensors
• Large amounts of data from social networks, mobile devices and sensors
6. Guiding research ques:ons
Is it possible to automa:cally collect, combine and analyze data from sensors
(e.g. environmental sensors) and crowd-‐sourcing (e.g. using mobile devices)?
Can this data be stored and processed, in order to extract useful informa:on
to aid planning and decision-‐making?
This leads to:
(i) What is the most effec:ve way to integrate and organize mul:ple
heterogeneous, autonomous sub-‐systems and sensors data?
(ii) How can data mining techniques be used to provide `smart’ outputs for urban
planners, architects and designers when proposing small interven:ons?
7. Computing!
Urban Planning !
Architecture!
Middleware
data collection
data integration
data analysis
Social Network
Twitter
Facebook*
Weather Station
Arduino
Crawler
Other
NFC
GPS Tracking
MSD Analysis [space]
Behaviour Analysis [people]
Survey/Interview
Media
Video
Image
Living !
Campus!
9. Middleware
• Middleware refers to the software that is common to multiple applications
and builds on the network transport services to enable ready development
of new applications and network services.
27. Conclusions and Future Work
• The Device Nimbus middleware can be used to collect/combine data from
heterogeneous sources.
• Device Nimbus can be used to build a richer understanding of urban systems,
based on data collected, leading to improved tools for planning and
policymaking.
• The full implementa:on of Device Nimbus will provide the means to effec:vely
monitor users’ rou:nes – help us to understand the use of small open spaces,
providing important feedback of collec:ve experience.
• We also plan to scale-‐up our ini:al inves:ga:on to include data collec:on from a
diverse range of loca:ons distributed across the main university campus.
28. Distributed and heterogeneous
data analysis for smart urban
planning
Eduardo Oliveira –
eduardo.oliveira@unimelb.edu.au
Michael Kirley
Tom Kvan
Justyna Karakiewic
Carlos Vaz