IEEE Future Directions
Related Articles
That the telecommunications network can be used as a gigantic, and pervasive, sensor is not a news. As an example the mobile network has to track the movement (and position) of cellphones to be able to connect them. Get those data and you can have a very accurate map of traffic in a city, analyse those data and you can “learn” the habits of a population down to a single individual and then you can predict what might happen next.
Several years ago (and technology has evolved enormously since that time) Telecom Italia, Ericsson and MIT’s Senseable cities worked together to show the potential of the mobile network to provide the pulse of Rome in Italy. The trial was made in such a way to protect the privacy of everybody, yet even through the use of anonymised data it was possible to see what was going on in the city.
Technology can be used to analyse the tiny variation of a WiFi electromagnetic field to detect presence and movement of people inside a home from the outside. Artificial intelligence can be used to analyse the field variation.
Now an article on Wired is reporting a study made by researchers at Penn State University showing how fibre networks can be used to detect movement of cars and people happening to drive/walk in the vicinity of the fibre.
Fibre networks are (mostly) running underground, a yard or less below the tarmac. Driving and walking generates tiny vibrations that in turns alter the propagation of the laser signals flowing in the fibre. The alteration does not affect the signal (the message coded) but can be detected and analysed using artificial intelligence (spiced with machine learning to “learn” the specific characteristics of the effect in that portion of the network).
The study showed that people walking create vibration in the range of 1-5Hertz whilst cars generate vibrations in the 50 Hertz range. Heavy machinery generate vibrations in the 100 Hz range.
What is amazing from the study is the resolution that can be achieved: alterations can be localised with a precision of 2 meters (6 feet and a half). It is like having thousands of sensors picking up activities above ground.
A city can use these data to monitor traffic, a factory can use them to monitor the operation of machines (creating a digital signature) detecting anomalies as they happen.
Interesting stuff!
