Auto-learn Light Levels

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • Andrew Millne
    Automated Home Ninja
    • Nov 2007
    • 269

    Auto-learn Light Levels

    What are people's thoughts on an "Auto-learn light levels" feature in Cortex?

    I've been tweaking my light levels today and was thinking that using the difference in light levels between when the light is off and when it is switched on would be a good way of learning an optimum light level trigger. My thinking is there will be a sweet spot where turning on the light will increase the light level just enough to make a difference but during the brighter periods of the day relatively no difference at all.
    Visit my blog and Follow me on Twitter...
  • Paul_B
    Automated Home Legend
    • Jul 2006
    • 608

    #2
    Interesting thread. I've thought about this based on time of year as well. During the winter when outside light levels are low then you can cope with lower light levels indoors as well. The reverse is also when summer sunshine requires lights to come on earlier.

    Of course this is just my observation

    Paul

    Comment

    • Gumby
      Moderator
      • May 2004
      • 437

      #3
      Isn't this going to be rather dependant on relative position of sensor, natural light, artificial light and the position most in need of light (due to particular task performed)?

      I guess you could use user demand (in terms of override/toggle requests) to help bias the system. But that raises a lot of other issues...
      ----------------------
      www.gumbrell.com

      Comment

      • chris_j_hunter
        Automated Home Legend
        • Dec 2007
        • 1713

        #4
        yep, interesting ... suspect same goes for heating - ie: in Winter indoors can be cooler, in Summer indoors can be warmer ...
        Our self-build - going further with HA...

        Comment

        • Gumby
          Moderator
          • May 2004
          • 437

          #5
          Originally posted by Paul_B View Post
          Interesting thread. I've thought about this based on time of year as well. During the winter when outside light levels are low then you can cope with lower light levels indoors as well. The reverse is also when summer sunshine requires lights to come on earlier.

          Of course this is just my observation

          Paul
          Kinda counter intuitive, since I would have thought the light needed for a particular task would not be seasonal, and if anything light would be needed sooner in winter due to reduced contribution from natural daylight.

          Perhaps this is most relevant in entrance areas (or anywhere you might be wearing photosensitive specs)? Or due you think this is primarily a psychological factor, similar to the need for heating boost when it's cold outside?
          ----------------------
          www.gumbrell.com

          Comment

          • Paul_B
            Automated Home Legend
            • Jul 2006
            • 608

            #6
            Originally posted by Gumby View Post
            Or due you think this is primarily a psychological factor, similar to the need for heating boost when it's cold outside?
            My own experience suggests it is psychological and very similar to the heating boost. Just like climate control in the car, in winter the temp is set at 21 degrees in summer I set it around 18 degrees.

            I guess it is quite a lot more complicated then first impressions may have you believe

            Comment

            • chris_j_hunter
              Automated Home Legend
              • Dec 2007
              • 1713

              #7
              another factor would be age etc of people - grandma would almost certainly like it warmer & lighter ...

              so looks like there's lots to go at with the tools Cortex provides !

              Maybe with a Mir:ror & an RFID doll for each person, so they can tell Cortex when they're home, and it can set /do things accordingly - eg; reading lamp brightness according to who & when (day, month, year) ... ?
              Our self-build - going further with HA...

              Comment

              • toscal
                Moderator
                • Oct 2005
                • 2061

                #8
                A lot more offices and museums are doing this. Using daylight to its maximum advantage can save on lighting costs. And more and more architects are considering lighting in the over all designs of buildings now.
                But it does produce a few headaches in planning.
                The main problem is the sensor placement, and control logic. Wrong sensor placement can lead to lights not coming on soon enough or too early and couple this with poorly designed logic control and you've got lights oscillating on and off. Another problem is type of sensor and control type. I know some systems just use a basic dawn dusk type controller and a relay. While others use actual light level sensors and some form of computerised controller.
                Then as already stated depending on the task will also depend on light levels. There are actual light level guide lines for various situations, if you want I can post them here as a PDF file. Then the older you get the more light you need. But this is where it gets really tricky as the colour of the light also plays an important part, lower lights levels are possible with certain colours, can't remember which colours or colour temps work better.
                IF YOU CAN'T FIX IT WITH A HAMMER, YOU'VE GOT AN ELECTRICAL PROBLEM.
                Renovation Spain Blog

                Comment

                • Karam
                  Automated Home Legend
                  • Mar 2005
                  • 863

                  #9
                  IDRATEK light level sensor equiped modules provide level information rather than a threshold switch, however within Cortex ON/OFF lighting control is based on threshold values. Variable lighting (dimmers) can be mapped via a user defined curve directly to a sensor value - but that's another story. In my opinion it is often important to use sensors in the actual area being controlled rather than using an external sensor since external conditions can be a crude and sometimes wrong indicator for specific internal conditions.

                  'Good' positioning of a sensor is of course important since although some diffusion occurs, the sensor is biased by what it is looking at (its colour, and whether its a light emitter being just some examples). For this reason, in a domestic environment setting, absolute light levels are generally less useful than setting thresholds to suit the particular sensor. I think in a more rigidly structured and legislated environment, such as an office, there will be a greater need for absolute light levels to be measured and controlled at the work position, but in such an environment it will also probably be easier to ensure more consistent sensor positioning and lighting elements for example.

                  Light oscillation and other erratic behaviour could indeed potentially be a problem in such a level based feedback system but it is prevented by the integration environment, dynamic signal filtering, and some hysteresis. For example Cortex knows when a light has been switched on and therefore that there will be a subsequent rise in light level, and there is often also some help from nature - unless you are pointing directly at a light, daylight tends to be brighter than artificial lighting.

                  The various sensor thresholds presently need to be selected by experience as suggested. The idea of automating this aspect and perhaps too seasonal variations is interesting but I'm guessing probably not so easy to do - not because of the technology but because of the more usual challenge of trying to second guess human nature

                  Comment

                  Working...
                  X