Re: OTC and Timeseries...
Posted in 2008
Topics: General Discussion
Ian Michael Gumby wrote: > While OTC points out a neat hack using Timeseries... > Terry Ralph long ago thought of other applications where Timeseries > could be applied. > > What would really be cool is for the power company to do the > timeseries tracking for you. > In the areas where they have the computer controlled meters and the > ability to read the meters in real time, it would be theoretically > possible for the power company to track and store this information in > set time slices (1 a minute). > > Now that would be a much more impressive feat and one that could > potentially have some serious value. Ah well, we'll see soon enough. There is an alternate perspective that says that seeing how much your consumption is could inspire you to change your behaviour, which is even more useful. Because we're going to run out of electrons Real Soon Now (TM) ... > (Imagine being able to drill down and track where and when power > requirements surge so that they could plan and be more effective. > Especially if we are to expect to see 'plug in' cars hitting the road. > (A second meter for your garage so you can track the electricity you > use for running the car separate from running the house...) > > With respect to more commercial uses... imagine Siemens Medical > imagining or GE Medical using IDS to help capture MRI slices, Pet or > CT slices. Or in the capture and rendering of 3D ultrasound images. > (Someone wrote a DICOM datablade, but I'm sure GE and Siemens have > their own...) Could even be used in part of a digial angio system. > > Of course if you live in the UK or Chicago, using IDS and Timeseries > could be using in the CCTV capture systems. Sure. I'd love to see you store images in a timeseries. > A more mundane use would be in capturing data in water treatment > facilities. > > The point is that while the blog has a cute example of a hack, there's > much more potential opportunities that are not being exploited. (Yet > another sales/growth potential wasted...) Still, it's much easier to sit there and tell everyone else what they're doing wrong than to go out there yourself and sell it, eh? :o) -- Cheers, Obnoxio The Clown http://obotheclown.blogspot.com
On Dec 14, 1:38 pm, Obnoxio The Clown <obno...@serendipita.com> wrote:
> Ian Michael Gumby wrote:
> Ah well, we'll see soon enough. There is an alternate perspective that
> says that seeing how much your consumption is could inspire you to
> change your behaviour, which is even more useful.
>
> Because we're going to run out of electrons Real Soon Now (TM) ...
>
Nonsense. We've got plenty of electrons. ;-)
The point was that while one person tracking one meter in real time is
a nice hack. Tracking literally hundreds / thousands of meters in
"real time" is a better hack.
(Of course you need to have a new fangled digital meter to make this
work. Something that it will take a while to catch on...)
You want to be more efficient, its one thing to change the behaviour
of your usage, but its another to be able to make the grid more
efficient. Also avoid things like brown outs or blackouts because of a
problem that happened elsewhere in the grid.
> Sure. I'd love to see you store images in a timeseries.
>
Are you that silly?
Look the short answer is that you don't.
The longer answer... if you can't figure out how to do it, then leave
it to the professionals and stick to juggling. :-P
(Try storing a reference to the slice in the timeseries and the slice
is stored in a blob space. )
> > A more mundane use would be in capturing data in water treatment
> > facilities.
>
> > The point is that while the blog has a cute example of a hack, there's
> > much more potential opportunities that are not being exploited. (Yet
> > another sales/growth potential wasted...)
>
> Still, it's much easier to sit there and tell everyone else what they're
> doing wrong than to go out there yourself and sell it, eh? :o)
>
Well, if I had a research grant and I was working in conjunction with
GE Medical or Siemens medical, then I could. If I could get an NSF
grant and a hospital willing to take one of their MRI's offline forpatient use and allow for some R&D, maybe I could do it. Oh wait.
Hospitals spend millions of dollars just to build a facility to house
an MRI unit or units. Then there's the price of each unit, and the
lost revenue from having the unit down for patient use.
(Try coming to Chicago for RSNA and check out InfoRad then you'd get a
better understanding why *I* can't do this as an individual.)
Now if *I* worked for IBM's IM in the OEM/VAR support staff, I might
find it interesting to find out who supports GE Medical or Siemens
Medical and talk to them about the potential of exploiting TimeSeries
in their products. But *I* don't work for IBM and *I* don't have
access to Blue Pages or the Siebel system that tracks all sales
opportunities to find out which git supports one of these clients. (Or
if I did work for the right department, I'd ask my manager or a co-
worker who I should contact.)
With respect to capturing data from a water treatment facility...
thats actually already being done. I wrote the OS and the first
downloaded program for the 6809 controller used by Nalco Chemical.
Then the consultants after me modified the program to capture and
report back the results so that they can be captured in to an Oracle
database. Their only instruction... don't fsck with the OS and the PID
calculations. This has been going on for almost 20 years now. The only
thing is that because your sample rate is something like once every 3
minutes or so, you don't get a huge advantage of time series. If
you're printing out daily or weekly graphs, you can query the
database, get the data "fast" enough to make you chart for your
report.
As to doing something like this on my own... First the market for such
a device is small. Nalco used to give this device away for free as
part of their contracts to supply the chemicals. Not to mention that
amount of testing and certification you need to go through ... not
worth it.
Like I said, these are some ideas that could be exploited by larger
companies already in the business of providing said products, not
something you can do on your own.
But hey! What do I know? ;-)
-G