A Conceptual Model and the Supporting Middleware for Composing Ubiquitous Computing Applications

Abstract

Given the resulting complexity of the ambient applications that one can form in the Ubiquitous or Pervasive Computing domain it is required to abstract the intricacies of a heterogeneous supporting environment (e.g., intrinsic characteristics of specific communication models) away from the application logic. These applications will be characterized by the increasing ubiquity of interactions between many possibly heterogeneous artifacts and services. This paper presents the Plug/Synapse abstraction, which provides a conceptual model for building ubiquitous computing applications in a high-level programming manner. GAS-OS is the software layer that implements the Plug/Synapse model and the concepts encapsulated in GAS, a generic architectural style, which can be used to describe everyday environments populated with computational artifacts. The paper examines also the design and architecture of GAS-OS, which is the minimum set of modules and functionalities that every device must afford, in order to be a ubiquitous computing artifact and participate in artifact collections.

Semi-tacit Adaptation of Intelligent Environments

“…In the former mode, people configure spheres by explicitly composing artefact affordances, based on the visualized descriptions of the artefact properties, capabilities and services [4]. In a highly dynamic system, such as an AS, explicit configuration is useful only for setting up the initial values or to model a default profile of the AS (see Fig.…”

Using ubiquitous computing technology to realise scalable intelligent agricultural environments

“…The work presented in this paper extends the scope of this research and applies ubiquitous computing technology to support resource management in a plantation, which is considered as a collection of individual plants. In this case, the approach of turning each plant into an ePlant and then use the plug/synapse model [9] to achieve collective resource management does not scale well, due to the large number of resulting interactions among ePlants and artefacts. Since considering each plant individually does not yield great benefit, we propose to group plants into ePlantations and enable each ePlantation to communicate as an individual with artefacts.…”

“…Τhe deployment of mixed societies composed of plants and artefacts is supported by ePlantation-OS which is a target application modification of the GAS-OS [9]. The ePlantation-OS is additionally composed of the following modules: Plantation state diagnosis module: responsible for the local assessment of the state of each ePlantation.…”

“…A rule-based knowledge base is required for the reasoning process; Machine Learning(ML) module: This module runs a ML scheme that uses different ML techniques (clustering, classification) so as to produce more {plant or environment}-oriented rules that will benefit to the adaptation of the system and increase the precision of the system. Interaction module: it implements the interaction scheme between plants and artefacts, in the form of the Plug/synapse model [9]; The ontology manager is responsible for the interaction of the nodes in a mixed society of plants and artefacts under the definitions of terms and relationships found in the PLANTATION ontology and for the management of this ontology. Key knowledge encoded in the ontology is the characterization information that relates a plant's status to the measurements provided by one or more sensor systems in the form of given threshold values or range of values or rules.…”

Talking plant: integrating plants behavior with ambient intelligence

“…This middleware is an upgraded version of the middleware presented in Drossos et all (5). Thus it encapsulates the basic features of the kernel and incorporates new modules, specifically developed for managing plant behaviour.…”