Brian Pickering

Taming the interoperability challenges of complex IoT systems

The Internet of Things is characterised by extreme heterogeneity of communication protocols and data formats; hence ensuring diverse devices can interoperate with one another remains a significant challenge. Model-driven development and testing solutions have been proposed as methods to aid software developers achieve interoperability compliance in the face of this increasing complexity. However, current approaches often involve complicated and domain specific models (e.g. web services described by WSDL). In this paper, we explore a lightweight, middleware independent, model-driven development framework to help developers tame the challenges of composing IoT services that interoperate with one another. The framework is based upon two key contributions: i) patterns of interoperability behaviour, and ii) a software framework to monitor and reason about interoperability success or failure. We show using a case-study from the FI-WARE Future Internet Service domain that this interoperability framework can support non-expert developers address interoperability challenges. We also deployed tools built atop the framework and made them available in the XIFI large-scale FI-PPP test environment.

Forecasting impact of technology developed in R&D projects: The FITMAN approach

A typical two-or three-year research project has an impact that is only really visible after the project has come to an end, at a time when there are no resources to monitor that impact. As a consequence, projects need to estimate/predict their future impact before they end. In this paper we describe the impact activity monitoring method in the FITMAN project. This method addresses the problem by accounting for actions to raise impact during a project and the planning for such actions after a project has ended. We also describe the socio-economic impact assessment methodology created in FITMAN, showing how this links to the impact activity monitoring method. Key to both is the assessment and monitoring of impact in three different areas: industry, society and the scientific community. Each area represents different challenges and we discuss their relative value to the overall assessment. We also report on our early experiences of applying this to ten industry-led use case trials in the FITMAN project. The insights gained by applying these methodologies can be more widely applied across domains related to technology management.