SMWCon Fall 2015/Semantic MediaWiki for Context Management in a Web of Things Use Case of Ambient Assisted Living

Motivation
The demographic change has originated the last years a new field of research and technologies, called Ambient Assisted Living. Ambient Assisted Living –shorthand AAL- comprises approaches and methodologies to offer elderly and impaired people an independent and self-determined life in their own living environment. AAL has the goal to support these people for mastering their life and to minimize the dependency to other people. But to achieve an appropriate support of the target group, there are some challenges which have to be solved. But first it has to be clarified what an AAL environment and application consists of. The main approach of AAL is to integrate different kind of sensors in the (living) environment of the end-user. These sensors can differ in usage, characteristics and functionality. There are stationary sensors which are integrated fix in the environment and there are wearable devices which are measuring at the user for instance the vital data as well as maybe the current location of the user. Ambient Assisted Living has the requirement that sensors have to be integrated in an unobtrusive way so that the user will not be bothered by the technology. Besides, there is the need to configure or adapt an AAL system to the user needs and characteristics. The target group has for instance mostly impairments which have to be compensated by the system. This implies that the whole AAL system has to adapt itself to the end-user. To achieve an adaption to the end-user, the environment of the user as well as the user characteristics has to be mapped into a domain model which can be dynamically processed by the AAL system. To adapt the services of an AAL environment, it is also necessary to supply user profiles and service specifications describing the capabilities and needs of the services itself for processing the appropriate task to provide the necessary help. Before an AAL system can be launched, all sensors and user profiles have to be created in an initial step. Besides it is a mechanism needed which can dynamically add or update ontology instances during the runtime of the system.

Challenges
One challenge is that mostly devices from different manufacturers have different protocols so that there is no standardized way to integrate different devices into an AAL environment. This implies the need of an abstraction layer to make it possible to integrate different kind of devices with a detailed description of their properties, capabilities and events of interest. Especially the sensors and the related devices and inventory have to be abstract from their underlying implementation with supplying machine readable information about these devices. Every living environment is different like every person who is participating the system, is different. This implies that an adaptive, context-aware and distributed AAL platform has to adapt itself to every environment, its changes during runtime and the user’s needs and actions in real-time. To make this possible it is essential to integrate, configure and -if needed- to update the AAL environment for individual purposes. It has to be considered that the integration of AAL sensors and platforms is a task that currently just can be done by technical experts who have enough knowledge about the sensors and their usage. End-users who are not technically minded have currently not the opportunity to adapt the AAL environment to their needs. Also in this case Semantic MediaWiki supplies helpfull tools for solving this issue.

Approach
The approach in this abstract is to create a light-weight ontology covering the relevant information of the AAL domain and to use as mean Semantic MediaWiki to provide a way to integrate different kind of sensors and also other relevant information about the AAL domain like user profiles, descriptions about the environment and participating services. Semantic MediaWiki generates -in the background- out of annotated pages RDF files. This RDF representation can be used to provide abstracted information about all relevant AAL instances. Besides, Semantic MediaWiki provides templates and forms to allow also unknowing users the creation of semantically annotated information in a Semantic MediaWiki. For every integrated device and service is information provided. So if a service needs further information about a device or another service and its characteristics and capabilities it can request the appropriate RDF representation of the appropriate device or service from the SMW or a connected triple store. This addresses a new approach called Web of Things which stands in relation to the Internet of Things. The Web of Things can be considered as an extension of the Internet of Things. Semantic MediaWiki can give one opportunity to model devices and to provide a description about their characteristics to participating services and applications. Our approach is to create the whole AAL ontology instances with Semantic MediaWiki and to provide these data for introducing them to the context mangagement system, which can generate out of these instances new knowledge that can be used by services. The realisation of these approach is foreseen for the AICASys project which has the goal to provide impaired people different kind of assistance appropriate to their disabilities. The AICA system is a distributed system which consists of different kind of (wearable and stationary)devices, a home automation system for integrating the devices at a low level protocol layer, different kind of services related to different use cases and a context management platform which gathers sensor data that has to be annotated in a semantically way. During runtime of the AICASystem new added ontology instances are synchronized with the context manager component so that just in time new information about environmental changes is available. All components are communicating via Web protocols like HTTP, REST or Websockets and are using JSON-LD as way to provide semantically enriched information about all participating things of the AAL environment.

Topics
The following aspects will be considered in this talk:


 * How can be provided to every user a simple way to integrate semantically annotated things of the AAL domain? What means can SMW provide to adress this issue?
 * How can semantical context data just in time be provided to every participating component?
 * How can be used in the AAL domain the Web of Things technologies and which role can play SMW?