I'm posting this as this question has come up in the past in other forums and I wanted to post the answer here:
Can the the TDC1000 and TOF measurements be used to uniquely identify a fluid?
Fluid identification works by measuring the TOF (TimeOfFlight) over a fixed distance which yields the speed of sound of the material. So normally FLUID ID works by characterizing the speed of sound of a limited number of fluids options that can be present and differentiating between them.
The trick is each fluid needs to be characterized over temperature as the speed of sound changes over temperature and they would need to select fluids that speeds of sound didn’t overlap otherwise they can't be uniquely identified.
I like to describe Ultrasonic Fluid Id as it being really good at detecting if the material present isn’t what’s supposed to be present.
That being the case for a consumer device there could be an infinite number of liquids and liquid combinations over temperature that could create similar speeds of sound and thus make it impossible to uniquely determine the fluid being measured.
However, if they used another physical parameter (ie: conductivity) they could amass a huge database after testing all combinations of liquids that could potentially identify unknown fluids.
For more on this topic refer to my app note: http://www.ti.com/lit/an/snaa265/snaa265.pdf
