Smart-Its: An Embedded Platform for Smart Objects
Michael Beigl
Telecooperation Office (TecO)
University of Karslruhe, Germany
Hans Gellersen
Computing Department
Lancaster University, U.K.
Abstract
Computer and communication hardware has become so small
and inexpensive to consider their embedding in everyday
objects. As a consequence it is expected that networked smart
objects will give rise to new types of application and in
particular such that are more tightly coupled with activity in
the physical world. Recent years have seen many design
examples for smart objects but for lack of a suitable platform
it has remained difficult to study applications that involve
networked operation of a diversity of such objects. In the
Smart-Its project we are addressing this concern with
development of embedded technology designed for post hoc
augmentation of everyday objects. This paper describes our
technology concept and the design of the 1G Smart-Its
hardware/software platform.
1. Introduction
The transition from mainframe to personal computing was
marked by ‘human integration’, considering human users no
longer as peripheral but as integral in computer applications.
Similarly, the current transition from personal to ubiquitous
computing is marked by ‘physical integration’, considering the
physical world around computer and user as integral part of
the overall system. A far-reaching approach to achieve such
physical integration is to embed computing into the objects
and artefacts that are subject of everyday activity. A hindrance
though is the lack of a suitable platform that would support
such augmentation of everyday objects.
Artefacts are commonly defined as ‘something created by
human for a practical purpose’ and it is compelling to build
on these familiar purposes while enabling new applications on
the basis of embedded computing and communication.
Artefacts thus augmented become smart objects that can be
tied directly into software processes to overcome the media
break between physical flow of activity and related flow of
information. The possibilities appear to be open-ended. For
instance in an enterprise environment, smart objects may
provide continuous access to their physical state and context,
or even embody autonomous behaviour in support of more
decentralized enterprise management. Likewise in emerging
interactive environments, smart objects may embody physical
I/O to be enabled as tangible user interface objects that
facilitate richer interactions between people and their
environments.
Miniaturization of components is presently reaching a
stage at which it becomes practical and affordable to embed
processing, networking and physical interaction into even the
most mundane objects [1]. This has inspired a range of design
examples built over the last years to explore application
opportunities and technology design challenges, for instance
the Mediacup (a coffee-cup that autonomously computes its
use context from embedded sensors, and serves itself to
potential applications in the local environment [2]), the Strata
Drawer (a chest of drawers that tracks its physical contents to
provide new forms of user interaction [3]), and the Pin&Play
noticeboard (a board that has smart pushpins autonomously
asserting priorities to visually alert users [4]). These examples
are generally one-off prototypes and can only provide very
limited insight into applications and challenges that may
emerge with more pervasive networking of smart objects.
Investigation of applications that involve a larger number and
diversity of smart objects has so far been hindered by the lack
of a suitable hardware/software platform. The generic
technologies available for augmentation of common objects
are presently limited to passive provision of digital identity
[5,6].
In this paper we describe the development of a new
platform specifically designed for augmentation of everyday
objects. At the core of our contribution are Smart-Its. These
are small computing devices ultimately envisioned to have a
form factor of a label that can be attached post hoc and
unobtrusively to arbitrary physical objects, to empower these
with processing, context-awareness and communication. We
approach this vision with development of a family of small
embedded devices that enable prototyping with different
configurations of wireless network, sensors and actuators, and
that support customization to particular physical objects
through software abstractions and development tools. The
different types of Smart-Its developed to date include a first
generation (1G) optimized for low energy-consumption and
deployment in longitudinal studies, a 2G device developed for
integration with the Bluetooth world1, and a “do-it-yourself”
device for very rapid prototyping and ease of modification
[7]. In this paper we focus on the development of the 1G
platform. We briefly reflect on the general Smart-Its
architecture and requirements followed by description of the
1G hardware, system software and application development
support.
2. The Smart-Its Concept and Architecture
2.1. Technology Concept and Architecture
Smart-Its are generic computing devices however it is
fundamental to our concept that we consider them not stand-
alone but always as secondary to a physical object of primary
interest in the application world. This follows the philosophy
of the “Disappearing Computer” which places computing in
the background of people’s interaction with their physical and
social environment [8].
1 Developed by project partners at ETH Zurich
