sensor array global sensitivity

[weeb0]

Hello,

I'm facing a big dilemma and I cannot find the answer by myself.

I have an array of different sensor. To ease the computation I plan to use the sum, the mean of
those sensors or use root mean square or some other method to group the sensor response in one
global.

By doing that, I wonder what could the the total sensitivity of the system ?! Is it physically possible
the total sensitivity will be lower than the sensitivity of the best sensor ? How to compute that ?

Thank you a lot!

My head is about to explode !!

 

[BradtheRad]

If the sensors all gave the same reading your job would be easier.

Sometimes you can't know beforehand what the data will tell you.

Taking simultaneous signals from multiple sensors is equivalent to moving one sensor to different locations and taking a signal. It opens you up to a greater range of readings. Is that what you mean by sensitivity?

You might also use a number of sensors to alert you to an anomalous reading in some area. Then you may wish to install additional sensors to give you more detailed data.

How to make sense of signals from many sensors? It's comparable to taking votes. The outliers may be 'rogue' voters, or they may be important. As to how to treat outlying readings, it depends on whether you wish to include all your sensor readings, or whether you intend to throw out the highest and lowest values as 'unreliable'.

Depends on whether your purpose is to lend greater legitimacy to your conclusion (the more data, the more weight)...

Or to average all readings (gather votes from sensors and treat them all equally)...

Or to have many simultaneous data (example, seismographic readings)...

Or to use one sensor as a reference (example, to discover that temperature inside an enclosure is always 10 degrees warmer than outside)...

Or to alert you to a problem (which smoke detector might be on fire).

Etc.