Internet of Things

About a week ago, close to 300 IoT champions from across Europe and overseas gathered in Venice for the annual IoT Week and the second International IoT Forum meeting. The venue was la Scuola Grande San Giovanni, and Venice was steaming hot due to the weather phenomenon called Scipione that brought hot winds from Africa to across Italy. So there was not only plenty of water flowing in the canals, but also into the throats of the participants.

The week was a successful assembly of the European research and innovation community coming from both academia, research institutes as well as smaller and larger enterprises and corporations. There were also officials from the European Commission present to talk about the importance of IoT for the European society, citizens and businesses. Keynotes were given by a number of prominent speakers and a lot of interesting research was shared, discussed and demonstrated. Topics ranged from technology to business and societal challenges like privacy, security and ethics. Technology covered everything from the default (but rather boring) IPv6 debate via IoT architectures, RFID, sensors and actuators, to cognitive technologies, AI, open data and the vast spectrum of IoT applications.

Adam Greenfield, founder and managing director of Urbanscale, gave a very interesting keynote on the topic of Smart Cities. He trashed the use of top down "optimum control theory" in city planning and advocated spontaneous order to grow from below. He referenced several Smart City proposals and visions by prominent global companies and said that they are all useless. When people claim that smart cities are about addressing the goals of the city, he claimed that cities do not have goals - "What is the goal of Venice?"

Professor Ken Sakamura of Tokyo University let us in on the pioneering work done by him and his group at the Ubiquitous Networking Laboratory over the years. TRON, T-kernel, and the ucode system which is a technology independent way of providing identities to objects and places are examples of work that prof. Sakamura stands behind. Thousands of ucode tags are embedded in the Ginza district to support tourists and other visitors with information. Some 2 million smoke detectors equipped with ucode tags have also been distributed as another example.

Everybody with some interest in IoT knows about Pachube, now Cosm, and what Usman Haque has been doing to establish an open platform for IoT innovation. In his keynote, he was not entirely different in opinion from Adam Greenfield. He was in fundamental disagreement with the data-info-knowledge-wisdon view and how people wrongly take the flow of data and information as knowledge. Building control structures to do optimization and efficiency was in his view "bad language", and that how we (engineers, society, decision makers,…) are trying build controlled infrastructures for data and information today is how city-planners did high-rises in the 50s and 60s - bad! No, IoT should on the contrary be about emergence, crowdsourcing, individual grass root level collaboration and to make people feel they are part of urban life and not detached from e.g. a Smart City.

But, not only "big thoughts" and idealistic approaches were discussed. The week was also packed with reporting from a number of different projects and research being conducted as well as the European agenda looking forward. Most of the projects reported are also part of the so called Internet of Things European Research Cluster (IERC), which is the focal point for European research on IoT.

One interesting session was on Cognitive Technologies for IoT. Cognitive here means to add the smartness, intelligence and self behaviour capabilities to IoT, i.e. adding the "Smart" to "IoT". Autonomic systems shall not be confused with cognitive either, since the former has "hardcoded intelligence" whereas the latter has learning capabilities. Cognitive capabilities also serve two basic but separate needs, one being how they can interpret the needs of the applications, and the other how they can help optimize system resource use. From that perspective, examples given ranged from "virtual personal assistants" like in the Butlerproject, to dynamic frequency allocation in 6LoWPAN networks from the ebbitsproject. The projects present were those from the Activity Chain on Cognitive Technologies as part of the IERC.

Another session dealt with the rather big topic of establishing a common understanding of or even agreement on an IoT Architecture. The IoT-A project's Architecture Reference Model served as the blue print, and a number of projects were there to "confirm" the validity of the architecture. Quite an interesting setup. The IoT-A project has published an initial reference model for IoT that can be retrieved from their web page.

Yet another interesting session was arranged by the IERC Activity Chain on "Semantic Interoperability". How do we make data useful, and how do we make it possible to both share data with others, as well as make use of data that others provide? How do we make the vision of an Open Data market place possible? On this important topic, a manifesto/white paper is being produced. Hopefully, we can come back to this in a later post.

As mentioned, the International IoT Forum had its second meeting with sessions for its four different Working Groups on "Technology", "Economics", "Societal" and "Legislation and Governance". I was in the Technology session panel and gave a presentation on Embedded Technologies for IoT. The discussion that took place was somewhat falling back to the old IPv6 discussion, and the discussion on end-to-end IP or not. The debate on "IPv6 or not" is however of remote importance as the IoT value does not lie in the plumbing, but what you do with the data picked up from the sensors etc. The plumbing is merely a necessity, nothing more. Of more importance is the notion of "Embedded Web Services" as something getting closer to the true value in IoT (but still not really the real value as it is also part of the plumbing). Finally, some claims were made of gateways being an evil thing not needed if you go for IPv6, but I argue they are still needed for different purposes, e.g. to shield resource constrained devices from "ill behaving applications" or to even run localized services.

To conclude, the week was really dense with information. To find out more about all the different activities and projects on the European level, the IERC is a good starting point. The IERC also has a number (14 to be precise) of different Activity Chains covering different aspects of IoT. Finally, released during the week was "The Internet of Things 2012 - New Horizons", which is the 3rd edition of the IERC Book. It contains an updated Strategic Research Agenda for IoT as well as a set of papers giving an international perspective on IoT. Apart from the hardcover copy, it will shortly be available as a pdf document.