seems there is a technique recently developed called
distributed temperature sensing using optic fibres.
in this paper
https://www.google.com.au/url?sa=t&rct= ... yT0ftHjEzW
the authors say:
Our SWI technique is highly capable of measuring the spectral shift in the Rayleigh backscatter along an optical
fiber with high spatial resolution and spectral sensitivity. The linear relationship between strain or temperature and the spectral shift enables distributed sensing along any standard single-mode or gradient index multimode fiber with millimeter-range spatial resolution over tens of meters of fiber with strain and temperature resolution better than 1μe and 0.1 ºC. The instrument is robust and simple to use. In addition, as the measurement does not require specialty fiber or discrete elements, sensors are very economical.
This might be ideal for tesla battery module, as the resolution is down to mm level ( each cell temperature would be easily recognised) and it works over a length of 10m or so of optic fibre ( probably long enough for each battery module). It would allow to recognise early which cells are generating more heat and likely to fail. Although the tesla battery is not designed for easy replacement of individual cells it might be integrated into later iterations of the design.
Just a thought.