Just as the Internet radically reshaped society, the Internet of Things (IoT) will have an impact on all areas of human life: from our homes, vehicles, workplaces and factories, to our cities and towns, agriculture and healthcare systems. It will also affect all levels of society (individuals, companies and state-level), from urban to rural and the natural world beyond. This makes it essential to have a proper understanding of IoT and the challenges which relate to it. The primary aims of this document are to (i) determine the scope of IoT, its origins, current developments and perspectives, and (ii) identify the main societal, technical and scientific challenges linked to IoT. It seems inevitable that IoT will become increasingly omnipresent. Indeed, it is set to penetrate every aspect of all of our lives, connecting everything (billions of new heterogeneous machines communicating with each other) and measuring everything: from the collective action we take at a global level, right down to our smallest individual physiological signals, in real-time. This is a double-edged sword, in that it simultaneously gives people cause for hope (automation, ­optimisation, innovative new functionalities etc.) and cause for fear (surveillance, dependency, cyberattacks, etc.). Given the ever-evolving nature of the IoT, new challenges linked to privacy, transparency, security appear, while new civil and industrial responsibilities are starting to emerge. IoT is centred around an increasingly complex set of interlinked concepts and embedded technologies. At an industrial level, this growing complexity is making the idea of having full control over all components of IoT increasingly difficult, or even infeasible. However, as a society, we must get to grips with the technological foundations of IoT. One challenge for education will therefore be to gradually increase awareness of IoT, both in order to protect individuals’ sovereignty and free will, and to initiate the training of our future scientists and technicians. A public research institute such as Inria can contribute towards understanding and explaining the technological foundations of IoT, in addition to preserving sovereignty in Europe. IoT will inevitably increase dependency on certain types of embedded t ­ echno­logy. It is hence necessary to identify the new risks that entail, and to devise new strategies in order to take full advantage of IoT, while minimising these risks. Similarly to the situation in other domains where one must continually seek to preserve ethics without hindering innovation, creating a legal framework for IoT is both necessary and challenging. It nevertheless seems clear already that the best way of facing up to industrial giants or superpowers is to take action at the EU level, as shown by recent examples such as GDPR. Furthermore, given the growing influence of technological standards on society, playing an active role in the process of standardising IoT technology is essential. Open standards and open source – conceived as a common public good – will be pivotal for IoT, just as they have been for the Internet. Last but not least, massive use of IoT can help better capture and understand the environmental challenges we are ­currently facing – it is also expected IoT will help to mitigate these challenges. The goals in this context are not only to reduce the quantities of natural resources consumed by IoT (for production, deployment, maintenance and recycling). We must also aim to more accurately evaluate the overall net benefit of IoT on the environment, at a global level. This requires determining and subtracting IoT’s environmental costs from its (measured) benefits, which is currently a challenge. The growing impact of IoT underscores the importance of remaining at the cutting edge when it comes to scientific research and technological development. This document therefore aims to (i) highlight the wide range of research fields which are fundamental to IoT, and (ii) take stock of current and future research problems in each of these fields. A number of links are made throughout the document to contributions made by Inria. These contributions are, by their nature, diverse (basic and applied research, open source software, startup incubation) and concern the majority of research fields on which IoT is based.