OK so serious question -- utility companies tend to have high-accuracy, company-operated/maintained meters (that effectively measure the total use of a bunch of

OK so serious question -- utility companies tend to have high-accuracy, company-operated/maintained meters (that effectively measure the total use of a bunch of customers). They're meant for like, identifying non-malicious/technical losses but it's possible to use them to identify customers who are bypassing/tampering with their meters, as: total use (as measured by the trusted meter) must be equal to ∑ accuracy(customer ID) * reported use(customer ID) and assuming there aren't many cheating customers, and enough measurements (smartmeters make this easier, 15 minute slices is a lot better than 1 or 2 month slices), and assuming enough variation/independence between customers, it shouldn't be that hard to estimate the accuracy of each reporting meter. I'm wondering if this sort of balance check (i don't know the proper terminology, this isn't my field of expertise, if you know more, i would love to be corrected!) would have been sufficient to detect the sort of misreporting mentioned in this article. Full text of the actual published article is here btw: http://ieeexplore.ieee.org.sci-hub.bz/document/7866234/?reload=true http://ieeexplore.ieee.org.sci-hub.bz/document/7866234/?relo...