Aspects of Pervasive Information Management: an Account of the Green Move System Emanuele Panigati, Angelo Rauseo, Fabio A. Schreiber and Letizia Tanca Dipartimento di Elettronica e Informazione Politecnico di Milano, Via Ponzio, 34/5 - 20133 Milan, Italy Email: {panigati,rauseo,schreibe,tanca}@elet.polimi.it Abstract—The Green Move project aims at realizing a zero- emission-vehicle (ZEV) sharing service that also includes perva- sive information distribution. In this paper we discuss the use of context-aware techniques applied to data gathering, shared services, and information distribution; we also discuss how a context-aware approach applied to these tasks leads to the reduction of (noisy) information delivered to the users and to the personalized distribution of information. Privacy of data is also a main concern in the realization of the project, and a privacy-safe approach to information distribution and advertising is adopted. The project, grounded in many results on the use of context-awareness already published by the same authors, aims at building a real-life system based on them. Eventually, we briefly describe the rapid prototype produced and the approach employed so far for the realization of the full system. I. I NTRODUCTION Nowadays technologies enhance most aspects of everyday life. A technology which is able to seamlessly integrate in our way of living as a part of it becomes pervasive [15]: from biomedical monitoring – e.g. continuous health care – to automotive (e.g. self-driving or assisted-driving vehicles), large numbers of very small sensors and embedded systems participate in processes and data flows through the support of many different technologies. Such a large number of involved entities generate interesting issues about energy consumption, network connections, computation resources and, last but not least, data management. Huge amounts of data, coming from possibly large collections of participating entities, have to be collected, re-distributed and analyzed in a reasonable amount of time, in order to obtain useful and up-to-date information. Such a scenario is instantiated in the Green Move (GM) [1] project, whose aim is a zero-emission-vehicle (ZEV) sharing service for the city of Milan. In Green Move the core services are surrounded by a social-like platform to support users in a large urban context. The ZEV-sharing service provides four different service configurations, designed to meet different user-category requirements: condo-sharing for users who live in apartments and decide to share a (set of) vehicle(s) for daily usage (e.g. going to the supermarket, taking children to school, ...). This configuration is usually two-ways: the user returns the vehicle to the same place where he/she got it, typically the condo parking; firm-sharing for firms outsourcing their company vehicles to the GM sharing service. This configuration is usually two-ways like condo-sharing; world-of-services users use a GM vehicle to reach a point of interest – e.g. a museum, or a department store, which has an agreement with GM – offering dedicated services to GM customers (e.g. having the museum ticket charged on the GM monthly bill to skip the queue at the ticket office). This configuration is typically one-way: the user shall release the vehicle at a GM reserved parking nearby the aggregation point, any further usage will be independent: the user could reserve a different vehicle (or the same if it is available) for moving away after having enjoyed the service; generic users whose needs are not represented in any of the previous configurations. The GM system also aims at providing a complete user experience of core and accessory services, like integrated services offered by GM commercial partners in the city, service and traffic information and advertising based on users’ interests and GPS position. To fulfill such objectives we propose a context-aware approach to realize and manage situation-dependent services and support processing of data flows to extract interesting information accordingly. The ap- proach drives the data flows since its gathering phases, even from sensors, selectively retrieving data only in quantity and format useful according to the current context: e.g. driving downtown is different from driving in the suburbs, thus the user reasonably expects different information – like traffic density or the presence of restricted areas – and with different frequencies. With a growing number of vehicles and users, the amount of collected and exchanged data will make the efficiency of advanced services a critical issue: in this perspective, the use of selective and efficient context-aware data gathering processes, which filter the information on the basis of the context(s) of its acceptor(s), can certainly improve the effectiveness and scalability of the system. The paper is organized as follows: we present the data management subsystem of GM in more detail in Section II; a perspective about how context is modeled in our approach is presented in Section III and specific applications of the pro- posed approach in Section IV. A prototype of the core context- aware functionalities is described in Section V. Conclusions and future work are presented in Section VI. 2012 IEEE 15th International Conference on Computational Science and Engineering 978-0-7695-4914-9/12 $26.00 © 2012 IEEE DOI 10.1109/ICCSE.2012.93 648