University of Murcia Faculty of Computer Science Department of Information and Communications Engineering Spain How to Intelligently Make Sense of Real Data of Smart Cities: Practical Approach Antonio F. Skarmeta
Big Data as Key for Smart Cities 2
Big Data Applications A. Public tram service of Murcia City B. Smart campus of the University of Murcia 3
A. Public tram service of Murcia City 4
A. Public tram service of Murcia City Data analysis techniques: A trip-chaining method to deal with incomplete data Allows to re-construct trips as origin-destination tuples. Based on several asumptions (e.g. the origin of the next trip is the destination of the previous one ). A Complex Event Processing (CEP) based solution to timely extract the trips from the records Set of rules implementing the logic of the trip-chaining method. A fuzzy clustering algorithm to identify groups with similar features More general profiles of the public transport are extracted 5
A. Public tram service of Murcia City Results:378719 transactions from 23400 unique users, the solution was able to extract 110697 trips 6
A. Public tram service of Murcia City Results: 7
A. Public tram service of Murcia City Conclusions: 1. Experiments confirmed that most of the tram users are concentrate on 2 areas. As a consequence, most of the traffic flow was concentrated in the line segment between the inner city and the campuses. This was of great help to come up with more tempting promotional campaigns focusing on such type of young users. The frequency of trains moving along the inner city-campuses segment was modified by considering the start and end time of classes 8
A. Public tram service of Murcia City Conclusions: 2. They also confirmed the under-use of the line segment towards the shopping-mall areas 3. The formal discovery of the stations' load in terms of trips' origin and destination would allow the service provider and the city council to better plan the whole public transport service of the city. The more important stations might be considered as hub points where commuters can easily transfer from tram to another kinds of transport. Such an information could be also useful so as to forecast future infrastructure needs in each part of the tram line. 9
A. Public tram service of Murcia City Future work: Combine social-networking data and mobile sensing capabilities to enrich detected trips. Social network sities and smartphones sensors able to detect wide range of contextual feature of a person. - More customized urban-trip navigators - Better perception of the general trafficflow of a city 10
B. Smart Campus of the University of Murcia 11
Partners Funded by: European Union Total budget: 4,862,363 ; EU contribution: 3,286,144 Area of Activity: Framework Programme 7 ICT Objective 1.4 IoT (Smart Cities) Period: 1 st September 2013-31 st August 2016 12
SMARTIE Platform
B. Smart Campus of the University of Murcia 14
15
B. Smart Campus of the University of Murcia Cross-correlation between outdoor environmental conditions and indoor temperature 16
B. Smart Campus of the University of Murcia 17
18
19
the Bayesian NN model implemented is able to estimate the indoor temperature with a mean accuracy of 0.91 oc and a mean error deviation of 0.063 oc 20
21
B. Smart Campus of the University of Murcia Results: A mean energy saving of 29% meanwhile comfort preferences of occupants was satisfied in the 91% of the cases. 22
Future work Implement Big data techniques to infer people behaviour patterns inside buildings Include such patterns as input of the energy management of buildings Validate our solutions to buildings with different context 23
THANKS Antonio F. Skarmeta skarmeta@um.es