Cooperating Smart Everyday Objects - Exploiting Heterogeneity and 
Pervasiveness in Smart Environments 

Frank Siegemund Institute for Pervasive Computing, ETH Zurich 


Pervasive Computing envisions the seamless integration of computation into 
everyday environments in order to support people in their daily activities. 
Due to recent technological advances, the computing devices that can be used 
to enrich our environments with computation have become increasingly smaller 
and less obtrusive, so that it is possible to embed them into mundane everyday 
things. Such smart everyday objects have the ability to communicate with 
peers, perceive their environment through sensors, and provide context-aware 
services to nearby people. However, the severe resource restrictions and 
limited user interfaces of smart objects make it difficult for them to realize 
services on their own. Instead, in order to implement context-aware services 
and because of their limited user interfaces, smart objects need to interact 
with other computing devices to make use of these devices' sensors and 
input/output capabilities. In consequence, there is an urgent need to 
facilitate cooperation between computational entities in smart environments. 

This dissertation addresses the problem of cooperation from the perspective 
of smart everyday objects. Its main contribution is to illustrate how 
augmented items can make up for their limited resources by cooperating with 
nearby computing devices. For this purpose, we concentrate on two aspects of 
cooperation: interaction between smart objects and cooperation between smart 
objects and mobile user devices. 

The first part of this dissertation focuses on how smart objects can provide 
services in cooperation with other augmented artifacts. In the proposed 
solution, a programming abstraction facilitates the design of collaborative 
context-aware services for smart objects. This abstraction groups nodes into 
sets of cooperating objects that bundle their resources and appear to an 
application as a single entity. Based on this approach, we present a software 
framework for realizing cooperative services on smart objects. It consists of 
(1) a description language for context-aware services, (2) a context 
recognition layer for smart objects, (3) an infrastructure layer for 
distributing data among cooperating artifacts, and (4) a communication layer 
that adapts networking structures to correspond to the real-world environment 
of smart objects. To evaluate the proposed concepts, we present an 
implementation on an embedded sensor node platform, the BTnodes. 

The second part of this dissertation focuses on cooperation between smart 
objects and mobile user devices as an example of how augmented artifacts can 
benefit from the heterogeneity of smart environments. In this respect, our 
contribution is to identify recurring usage patterns that describe how smart 
objects can make use of handheld devices: (1) as mobile infrastructure access 
points, (2) as user interfaces, (3) as remote sensors, (4) as mobile storage 
media, (5) as remote resource providers, and (6) as user identifiers. We 
describe these usage patterns, report on experiences with prototypical 
implementations, and present several application scenarios illustrating the 
applicability of the proposed concepts.