Near field interactions with the internet of things

This paper focuses on annotating physical objects with contextual information from the viewpoints of affordance theory (Gibson 1979) and practical design of “Internet middleware”. Our starting point is the notion that linking user-generated meanings to everyday objects equips physical artifacts with new types of affordances, which we call here as social affordances (Kreijns &Kirschner 2001). These social affordances differ from the notion of perceived physical affordances (Norman 1988) in three essential aspects. First, instead of functional, their nature is principally social/socializing (Adler 2001). Second, instead of being defined by designers, social affordances are primarily defined by users. Third, unlike physical affordances, social affordances are historically accumulative; the more people annotate the same object, the richer its affordances become. The historical-accumulative nature of an annotated object is enabled by its positioning in a networked information ecosystem, from which judgements about the object may be gathered over time. Virtual-physical linkages such as bluetooth, RFIDs or printed labels connect these two worlds for the annotated object. In such an ecosystem, fragmented knowledge is linked by the essential actions of identification and pointing. In the Architecture of the World Wide Web Vol 11, URIs are described as "identifying distinct resources", and it is claimed that "global naming leads to global network effects". By identifying resources with URIs, references in data from multiple sources and services can be merged to form a rich aggregation of knowledge about a resource. The most valuable information about an object in an information network is not found in descriptions or traits of the object: it is in the relationships that the object holds with other resources. By pointing at things2 and saying why, datasets may be linked and the bigger picture may emerge. 1 http://www.w3.org/TR/webarch/ 2 http://www.plasticbag.org/archives/2005/04/the_age_of_pointatthings.shtml Following the success of Open Source, an Open Data movement is occurring online that seeks to gather, publish and enable the reuse of rich machine-readable datasets. This data would previously have been available only to large institutions having the financial means or the time to obtain it. By positioning an object’s information in relation to a shared dataset such as Wikipedia, Open Streetmap or Thinglink, (e.g. http:// thinglink.hackdiary.com/thingtagging/), the network effect is enabled and flat lists become rich graphs of information3. The necessary critical mass of annotation systems and annotated knowledge is not expected to emerge solely for the purposes of annotating physical objects. By bootstrapping on existing Internet infrastructure URIs, HTTP, and higher-level facilities such as blogging, locative services and social bookmarking applications object annotation can grow as the Web itself grows. For example, del.icio.us, the social bookmarking application, does not care what the URLs it annotates are pointed at. Specifying points in an information space whose axes are User/URL/Tag/Time, its applications emerge from use. The purpose of our paper is to discuss possible information architectures for social affordances in the context of a free open database called Thinglink, where people can register unique identifiers for meaningful objects. Central questions are: What kind of social affordances (relations between people and objects) are suggested in the descriptions of thinglinked objects on Thinglink database? What is the role of the network in supporting/expanding the identified social affordances? What Internet services (eg thinglink.org) are a natural fit with, or can be easily mapped to, this activity? How do these design considerations affect the building of web APIs or other 'internet middleware'?