I'm designing two brand new front ends for a scanning surface voltage tool for semiconductor manufacturing. Normally there are two ways you can measure surface potential—the vibrating Kelvin probe, which gets you the actual voltage but is very slow, or the fast-scan method, which is fast but gets you only the 1-D derivative of the potential.
Working with a leading semiconductor tool vendor, I've come up with both an improved transimpedance front end and (more interestingly) a new class of front end amplifiers that should be able to get the actual voltage like the Kelvin probe while going fast like the scanning method, while providing at least 15 dB SNR improvement over either one.
Update, March 3 2013: The TIA is done, and the new concept is underway.