Embedded networked sensor systems promise a revolution in monitoring and control. However, real-world examples are proving notoriously difficult to deploy and maintain, and as yet only relatively small scale successes have been published. This workshop will provide a forum to combine the experiences of international researchers, to create a systematic understanding of the challenges of building systems in the face of system and environmental instability, pointing the way towards a science of sensor systems software. In doing so we wish: To share experiences of real IoT developments “in the wild”. To understand the differences between the theoretical performance of results and real experiences. To map failure to causes and to explore the more and less obvious causalities. To produce a list of research findings and areas to prioritise for further investigation.