Sep 262014
 
 September 26, 2014  Court

As reported by John Wesley Hall of FourthAmendment.com:

A smart electric meter that transmits information about electric usage every 15 minutes is not a search and seizure. Naperville Smart Meter Awareness v. City of Naperville, 2014 U.S. Dist. LEXIS 134861 (N.D. Ill. September 25, 2014)*

Read an excerpt from the opinion on FourthAmendment.com.

 

  One Response to “A smart electric meter that transmits information about electric usage is not a search and seizure”

  1. That’s an interesting question the person who wrote the article made.

    “That it’s transmitted every 15 minutes and a meter reader doesn’t come once a month anymore is a search where the latter isn’t?”

    as an analytical chemist i’m supposed to be trained to be able to read chromatograms. Power usage over time is nothing more than a type of chromatogram.

    The data given in one big shot obscures who is home at what time. What was turned on. Any timers in the house that might be going to click on, or off.

    But when one plots power usage versus time, it’s no longer obscured. I can’t upload an image here, so I guess people can google chromatogram and look at one for themselves.

    So anyhow, as the time interval to upload the data decreases the picture of what you are using (and how you are using it) increases.

    A monthly reading doesn’t show much except the total used by a certain household.

    A 15 minute reading over time (days, weeks, months) will show a pretty accurate baseline. That baseline could be the blower motor on your gas furnace always on. When the gas furnace kicks in the power-Vs.-time chromatogram may show you have a variable speed furnace since the power consumption of the blower will increase.

    During night hours that baseline will return, what? it shows a total of a 60-watt over usage compared to daytimes when someone might not be home? Well we can conclude you have a minor increase for that one day which will show you left a light on in the house.

    Or, what could be deduced with fairly high accuracy if it showed a ~100-watt power draw every 15-minutes at night which is different between the night chromatogram and the day chromatogram when you are gone to work? I would deduce that person has a medical condition, like Apnea, and thus uses a CPAP machine.

    With a 15-minute sample time all sort of interesting, and private, details emerge over time.

    If they reduce that sample time to 5 minutes, or pretty much instant, then that baseline and everything else is so much more clearer.

    So back to this post now.

    Do I agree with it? Nope, not at all. Anyone here who is trained in reading chromatograms should be able to pick out the errors in that post.

    1. Data from the City’s smart meters shows only total usage and no further details than that.
    That is so wrong… It shows consumption versus time, versus time.

    2. Because there is no reasonable expectation of privacy in that data…
    Again, one could easily pick out a CPAP machine over time (a few days). A CPAP will stick out like a sore thumb actually since there will be routine to it (on and off times, and power usage, time of use).

    3. But that same guess could also be reasonably made by any member of the public walking by the residence who notices a car in the driveway or lights in the windows-that is not information that can be reasonably expected to remain private.

    flaw here is that it’s more than a car in a driveway, or a light on. It will show medical devices in use. CPAP is a perfect example of this.

    I don’t know if this would constitute unreasonable search and seizure (i’m just a simple chemist), but I know I could likely read this power chromatogram and build a picture. I don’t think it would be much different. Instead of ions, for example, I would use a toaster, microwave, various medical devices, types of light bulbs etc. As long as I have the baseline to subtract, and the approx wattage of diff type of instruments, a pretty good picture would come out I think.

    Think of it all like a chromatogram. It would be a little less precise, but that all depends on the sample times. I wonder if marketers would and could use this type of data? Just like other behavioural data. Do they?

    Hmm this post made me think a bit. I didn’t expect that. I also think the court made some errors/flaws in reasoning here. Maybe lack of knowledge. Someone bring that judge a chemist. Back to being a bum now. 😉

    or am I wrong?

Sorry, the comment form is closed at this time.