Friday, April 16, 2010

What the Internet of Things is not....

Tomas Sanchez Lopez from the Auto-ID Lab Cambridge writes about what the internet of things is not. From my point of view he has five good arguments...
  1. IoT should no longer free-ride on Weiser's ubicomp term - whereas there are similarities Weiser much more focuses on the user aspect and seamless integration instead of networked things
  2. IoT also goes beyond network protocols as it aims at integrating heterogeneous technologies.
  3. Whereas these technologies provide the basis for IoT they still do not ressemble the IoT, as to get the IoT working a range of services, e.g. discovery services, are necessary.
  4. The same argument applies for embedded devices, they augment things by sensing - but again it's the context of things these nodes are being used to make sense out them
  5. Applications are not the IoT, well, philosphically this should be right, but generally applications are always a nice way to communicate visions - they are actually necessary drive business and development.
I would still like to add social networks making sense of user's preferences and relationship to each other, a very important component to also expand the IoT beyond business corporations to everyday users. Concerning enablers, there is no way around mobile technologies, especially mobile phones, our daily companions. Yes, mobile phones are NOT the IoT but a major gateway to services and urban sensing.
What I actually would like to emphasize most are the business aspects. The entire hype about RFID and urban sensing using mobile phones was mainly driven by business opportunities. Companies saw a new era of gaining insights into their business processes and to optimize their supply-chains accordingly. Right, RFID adoption has not happened the way is has been predicted 10 years ago, but it still is probably the most widely deployed IoT technology. In the long run to get IoT out of the labs and out of the hands of technical visionaries, business needs and applications have to be identified in order to drive the development.
What could that be?
  1. see things you haven't seen before - visualize product flows, energy, carbon footprint, any kind of information bound to things and locations, e.g. using AR
  2. find things you might have forgotten - track & trace history, time series of sensor data
  3. share things you might have been too lazy before - crowdsourcing and mashups of user-generated data, comments/opinions to things, movement patterns
  4. finally, have even more precise information about what's going in order to act accordingly
Of course, if there wouldn't be challenges, it wouldn't be a research topic - hopefully, we can stimulate theses discussions at IoT2010 in Japan later this year...

No comments: