There is much to-do about RFID (Radio Frequency IDentification). My first project as a biomedical technician was to build a passive implantable pressure sensor that used the principles of passive RFID to measure the pressure within eyeballs/brains. This was in 1963 and the absence of the kind of convenient access to the lore of previous generations via Wikipedia and the Web in general left me naive about what I was doing.
Part of the fallout from that project was the technology of RIAS (Remote Infrared Audible Signage) commercially known as Talking Signs.
The subject of this post refers to "electro-magnetic identification" of which both these technologies are a subset.
In general, RFID is non-directional and rather limited in range whereas IRID is highly directional and virtually unlimited (theoretically) in range. The system known as PointLink would make use of the principles of IRID which has been mainly used as a mere presence/absence of IR sources rather than specific identifiable markers that could be used to present URIs for expanding information about them via Web use.
The before-long ubiquity of such systems presages a navigation/information system that could be the universal equivalent of what in the blindness field is "orientation/mobility" thus automating many of the tasks of getting about and learning about the environment heretofore only available through visual cues, particularly signs. Because they are just serial-numbered lights with known locations, they don't depend on language, etc. to be useful.
It's quite likely that their implementation will happen by individual placement of numbered transmitters by "Johnny Appleseed" methods since there is no need for them to be pre-catalogued, and anyone will be able to devise uses for such networks.