True self-configuration for the IoT

Abstract

For the Internet of Things to finally become a reality, obstacles on different levels need to be overcome. This is especially true for the upcoming challenge of leaving the domain of technical experts and scientists. Devices need to connect to the Internet and be able to offer services. They have to announce and describe these services in machine-understandable ways so that user-facing systems are able to find and utilize them. They have to learn about their physical surroundings, so that they can serve sensing or acting purposes without explicit configuration or programming. Finally, it must be possible to include IoT devices in complex systems that combine local and remote data, from different sources, in novel and surprising ways. We show how all of that is possible today. Our solution uses open standards and state-of-the art protocols to achieve this. It is based on 6LowPAN and CoAP for the communications part, semantic web technologies for meaningful data exchange, autonomous sensor correlation to learn about the environment, and software built around the Linked Data principles to be open for novel and unforeseen applications

IoT-O, a Core-Domain IoT Ontology to Represent Connected Devices Networks

“…This choice is motivated by the possibilities it opens: -Applied to data, it is a way to bring context-awareness, as presented in [10] -Applied to devices, it enables Thing discovery or self-configuration [11] -Applied to services it enables automatic composition as in [12] However, for concrete applications, the model should also by decidable, and in reasonable time, which de facto excludes an OWL-full model: OWL-DL is therefore the best choice. All surveyed ontologies are expressed in OWL-DL.…”

Context-Awareness in Task Automation Services by Distributed Event Processing

“…The employment of Complex Event Processing for Task Automation Services is a novel field of application, where only very first approaches have been published [5]. In general, related work shows task automation approaches conceived to solved particular problems, that lack of the flexibility and personalization capabilities that characterize TAS's.…”

“…Several authors propose systems where built-in rules are triggered by correlated events coming from different sensors [2,4,8]. SPITFIRE platform [5] is close to TAS's vision, since it provides a user interface to set up rules (called queries).…”

“…-Web-based TAS's such as Ifttt 3 , Zapier 4 and Elastic.io 5 are deployed as cloud services. They collect personal events by accessing appropriate web services on behalf of the user and provide a simple rule editor.…”

SEMIoTICS: Semantically Enhanced IoT-Enabled Intelligent Control Systems

“…Other works [19], [20] model IoT components as Web resources, thus enabling their online discovery and utilisation of their capabilities. An attempt to design selfconfigurable systems composed of IoT components has been also presented in [21]. There is also work from the "smart factory" domain and the Industry 4.0 concept, with teams working on modelling the information about factory objects (hardware or software) using semantic knowledge formalisms [22] and also modelling the IoT resources and processes towards optimal resource management [23].…”