[DigitallyFidget]

This isn't news to me at all. For ATM/Pin pads, I wipe my fingers across every button as I press in the code, so it obscures what was actually pressed, and linger fingers on keys not even in my pin code. With enough practice, it doesn't take more than a second or two longer than normally entering it.

As far as keyboards, I really don't ever interact with computers that don't belong to me or aren't in a secure area, but I have a custom scripted "keyboard" USB circuit thing that emulates keypresses for me. I don't know what to even call it, but it's like a mini Arduino sorta thing that emulates a generic Microsoft keyboard to whatever you plug it into. It looks like a stick of RAM with a USB plug, kinda. I have a few preset buttons that'll type in my login info to automate logging into things. I made it as a hardware password manager.

[marbu]

Indeed, this is not a new idea. The news here is about particular implementation for extracting passwords from QWERTY keyboards.

That said, this is not a big problem for ATM pin pads with metal keys, because these conduct heat well and so a heat pattern is hard to detect after few seconds. See: https://www.youtube.com/watch?v=PJCfTlQ82Fw

[woeh]

I've been in a hotel where the rooms had pin pads as locks that required you to press two random numbers every time you want to enter. The pad was a bit sensitive to fingerprints, but due to this mechanism there would be fingerprints all over the device.