Signature Detection and Matching

Suddenly dawned that a common theme in all of our products and services and research has been signature detection... whether it is very straightforward in Analytika, or a means for discovery in BACnet or Sunspec, or more nebulous in such detection for identification in Modbus (and other) mapping layouts.

Have been reminded of this by the similar methods we have seen security services companies like Genians use on port layouts to determine the type of system with which they are dealing - printer/PC/IoT, Windows/Linux, and others - and thence the types of issues to expect.

Sypris has new ways of detecting and matching hardware with a sort of silicon signature.

There have been many decades (since the 1980s and before) of physical sensing of signatures in building and industrial automation.

Motor and pump vibrations and audio signals.
Power line signals - Les Norford, MIT, et al.
Pressure line signals - Fred Schweppe, MIT, et al.

Such signature detection systems can be thought of as the genesis of monitoring via IoT.

This has been reborn recently with the likes of Zensi (and thence Belkin). Even RFID signatures can be thought of this way... not surprising they heralded the inception of IoT courtesy of Kevin Ashton (P&G), David Brock and Sanjay Sarma (MIT) and their kin.

Signature detection was not even new then. The grey beards had long been able to tell a motor or pump was failing simply by listening. Thermal signatures of switchgear, circuit breakers and bearings have long been used to predict their failures. RF signatures are used in RADAR and other systems. Maps and landmarks have always been an option for knowing where you are. The words themselves "signature detection" define the age old method of visually identifying and verifying documents (paintings, writings) and the trust we should endown them. Further witness the "fist" of a telepgraph operator. And the unique sounds of our voices and music.

Also in Cimetrics News