Ambient Amp: An open framework for dynamically augmenting legacy Websites with context-awareness

Abstract

Emerging context frameworks enable Websites to interact with the Internet of Things directly from the browser; however, Websites must be specifically designed to utilize such context framework support. As such, the majority of “legacy” Websites remains context-unaware. This paper presents Ambient Amp, an open framework for dynamically augmenting legacy Websites with context-awareness, without requiring browser extensions, proxies or Website reengineering. Amp provides an extensible Bookmarklet framework that serves as a conduit between the user's browser and a server-side repository of Amp plug-ins, which can be used to dynamically augment any 3rd party Website with new content, adapted behavior and context framework support - enabling augmented Websites to sense the user's physical environment and interact with discovered smart devices. This paper introduces the Amp architecture, its fully operational prototype and an example Amp plug-in that augments a well-known photo sharing Website with the ability to stream selected images to networked media devices discovered in the user's physical environment. We also present a preliminary performance evaluation, which indicates that Amp is suitable for deployment on many commodity mobile devices.