Context Watcher ─ Sharing context information in everyday life Johan Koolwaaij 1 , Anthony Tarlano 2 , Marko Luther 2 , Petteri Nurmi 3 , Bernd Mrohs 4 , Agathe Battestini 5 and Raju Vaidya 4 1 Telematica Instituut, PO Box 589, 7500 AN Enschede, The Netherlands, koolwaaij@telin.nl 2 DoCoMo Euro-Labs, Landsberger Strasse 312, 80687 Munich, Germany, {tarlano,luther}@docomolab-euro.com 3 Helsinki Inst. for Information Technology HIIT, PO Box 68, FI-00014 Helsinki, Finland, petteri.nurmi@cs.helsinki.fi 4 Fraunhofer Fokus, Kaiserin-Augusta-Alee 31, 10589 Berlin, Germany, {bernd.mrohs,raju.vaidya}@fokus.fraunhofer.de 5 Nokia Research Center, PO Box 407. FI-00045 Helsinki, Finland, agathe.battestini@nokia.com ABSTRACT We present the Context Watcher, a mobile application that enables mobile phone users to easily and unobtrusively share personal context data such as their location, heart rate, speed, or view, with their mutual consent. Not only can the data be shared, it can also be used as input for information services, to adapt applications to the context, or to automatically derive daily patterns and situational information, such as ‘meeting with a supervisor’, ‘is a regular visitor’ or ‘having the best condition of all your friends’. The application is built atop of the MobiLife context management framework, a generic approach enabling context discovery, exchange and reasoning. In the context management framework, different entities, context providers, are exposed to and interact via the internet. The context management framework and its principles will be discussed in detail, as well as the different context providers that play a role in the Context Watcher application. The Context Watcher has a large user base of 100+ users, some of which are 24x7 power-users taking the Context Watcher wherever they go, enabling the research on daily patterns and situations to work with real-life data, resulting in powerful algorithms and better understanding of context and context-awareness. KEY WORDS Context-aware computing, context reasoning, context tagging, mobile applications 1 Introduction When a mobile user will be late for a scheduled event, such as a dinner party at a friend’s house, the easiest way to resolve this unanticipated situation is to call their host, apologize for the delay, and provide additional information concerning their estimated arrival time. A drawback of this resolution is that this requires synchronous interaction between all parties, which causes the cooking party to leave the stove for a critical minute or to juggle the phone in one hand and the spoon in the other hand. It would be much easier if the host could check asynchronously where his guests were located, and get an alert in case they would not make it in time for the scheduled dinner time? With today’s current state of technology, scenarios such as this are not at all futuristic, but are possible to be implemented in your everyday life. Leveraging on these possibilities is exactly the goal of the IST project MobiLife [33], which brings recent advances in mobile applications towards users and group of users for making their everyday life easier. The main target user group of the MobiLife project is the family, especially families with members living in different locations, e.g., with children going to college. One of the project’s main technology aims is to design a general framework that can support the provisioning of applications and services that are capable of adapting their behavior and functionality to information that characterizes the situation of the user. In short, we refer to these types of applications as context- aware applications. MobiLife will develop and demonstrate context-aware applications that are easy to use, targeted to be used within the family group, and that satisfy real needs of the family members. The rest of this paper is structured as follows. Section 2 presents the MobiLife context management framework together with a short overview and specification of the interfaces and messages of the key components required for delivering context-aware functionality. Section 3 presents the Context Watcher application as an example of a context-aware application that is built atop the context management framework enabling family members to stay better informed about each other and their friends, without the need for synchronous interaction by using automatically derived context information. Section 4 presents highlights of context providers that are