Leveraging Machine Learning for IoT synergy
In its 4th iteration, this workshop will explore recent trends in convergence in IoT systems, with a special focus on heterogeneous data management. While mainstream IoT research largely exists in silos, we are building an international community of researchers who are focusing on interoperability between IoT systems, in developments that transcend single-function IoT deployments. While interoperability and functional convergence remain a pressing challenge, recent trends in multi-homing IoT architectures, gateways with multiple connectivity modes, and cataloguing systems that enable rapid resource discovery, have led many advancements in IoT interoperability. These developments are building on recent strides in managing hardware-agnostic low-power networks, which are promising many new frontiers in convergent operation.
We propose to focus the scope of C-IoT 2019 on topological remedies to handle Big Data communication and scalable IoT services. While many hurdles face synergistic IoT development, we will focus on techniques from Machine Learning, to aid IoT convergence on data and information planes. That is, as we are growing more able to communicate between heterogeneous IoT architectures, it is ever more pressing to address data compatibility and information extraction from heterogeneously sourced data. This includes challenges with data representation, meta-data tagging practices, establishing quality of resource (QoR) and quality of information (QoI) measures in heterogeneously sourced IoT data. More importantly, scaling such IoT systems is inherently tied with trusting such data, and our inference in deriving knowledge from data.
To this end, C-IoT 2019 topics will span architectures, frameworks and implementations that address IoT interoperability, across resource, data and information planes. That includes traditional IoT interoperability techniques (i.e. gateway based, hub-based, mixed-mode APs with heterogeneous IoT services, service-level cataloguing, etc), in addition to Machine Learning techniques that attempts to quantify QoI and QoR, as well as aid information extraction from IoT data. There is a clear demand in addressing IoT data management, and interoperability across large-scale IoT systems, and we wish to enable a dedicated venue for these pioneering directions.