X10 Sensors FAQ

Last Updated November 25th, 2005 - note comment at the bottom of the page ... ;-)

While the main focus has been on the webcam & webcontrol - webcam FAQ - I decided it might be fun to demonstrate a few more cool/geeky things. All of these are independant of the webcam ... and as with any "real-world" analog devices, the answers are not always "correct" and there is some "slop" in 'em, so you get some false positives & negatives. I'm don't know if I'll ever get around to documenting how all of this works in excruiciating detail, but you can always do a Web Search to find other folks who have! ;-)

This page is perhaps mis-labeled, since now (2004), most of these devices all communicate via Ethernet (wired or wireless) ... although the actual control of the lights are still done via X10 as-is the garage sensor.

Weather Sensors: Nothing special here - just gather a variety of weather data. We are a mile high, so the pressure is corrected to sea-level. As implied above, these tend to bounce around a bit (and sometimes don't even provide data) with the wind direction (understandably) being especially variable if the wind speed is close to zero. These were added in Halloween/2003.

Garage Sensor: Again, nothing fancy here, you've probably seen a similar device advertised (or at Open Houses) where a light comes on in your kitchen/bedroom if you left the garage open. Typically works off of a magnetic sensor. It also has the ability to tell you if the garage is actually in motion, so if you get "lucky", it will tell you this (plus you may see the garage door partway up in the picture, depending on the timing).

Ground-Motion-Cam: Standard motion detector type stuff. It has a little bit of a difficult time getting a correct reading at night ... but fortunately there is a bit of light from the bazillions of lights - or however many are currently ON! It is usually in the ballpark without too many false positives/negatives. If you look closely, you can see the person in the picture, *IF* the camera is pointed in that general direction, but this is often hard to to see unless they are right in front of a group of lights, plus there is a delay between when this looks for any motion and when the picture is snapped, as these are two seperate events.

Road-Cam: Similar to above, but looking at the road in front of the house. While a moving car is an easy object for the motion detector to pick up, it's again a challenge to get the field-of-view correct ... plus since the webcam picture is not taken at exactly the same time that the motion detector is queried, there may be a disconnect. Savvy folks might ask if you can estimate the speed of the cars based on the "smeer" of the lights and knowing the camera shutter speed ... I haven't done any speed/timing tests! ;-)

Sky-Motion-Cam: Now this is COOL!!!!! I live about 30 miles NW of Denver International Airport ... but more significantly, about 10 miles North of Jeffco GA airport, so there is a fair number of airplanes that fly overhead at reasonably low altitudes. Airplanes are always supposed to have red/green lights on the left/right wing, and they may have their "main" lights on, which is all visible to your eye, especially if they are oriented toward you. So I wondered if the motion detectors (with the gain turned as high as possible) could pick up a moving airplane. it turns out this is doable - note that stars are visible, but don't move much! So then I said to myself, could I determine WHERE the airplane is by restricting the field-of-view of the motion detectors - i.e. put 3 of 'em up there pointing to the left, middle, and right ... and it works pretty well, so depending on which one indicates motion, that is mentioned. BTW, as an aside, I noticed that the webcam DOES pick up the stars (look closely at the images), but their brightness appears to be variable and are sometimes not even visible. I guess this is just CCD randomness - there's just not a lot of light coming from those.

Sky (Cloud) Sensor: This was even tricker than the Sky-Motion-Cam described above, but it uses the same Cams. At the top of the hour, the images from the camera's are downloaded and analyzed, with an attempt made to see how many stars are visible against a "baseline" of a cloudless night. While the results are pretty variable, an assessment can be done of how many stars, depending on how many are visible, that gives some idea of how cloudy it is.

Total house amperage load: This is fairly easy - just a MEASUREMENT of how much total amperage is coming through the house breaker box. This includes other stuff besides the halloween/xmas lights - note that in the "lights box" I show the total amps from (only) halloween/xmas lights currently ON, but this is CALCULATED based on what zones are currently active - i.e. while I did take one-time readings of the amperage/circuit, there is NOT an an active amperage measurement per circuit, only of the entire house. BTW, LED Christmas lights will reduce your amperage load considerably, but they are still pretty pricey.

Steady-State Voltage: If you have ever measured voltage over a period of time at your house, you'll see that it varies quite a bit; another good reason to have Surge protectors, etc. And as you can imagine, switching bazillions of lights ON & OFF can cause quite a bit of fluctuations, so we take a measurement after an ON/OFF command, and if there is a sag/surge, that is shown. This "Steady-State Voltage" is based on a measurement taken at the very end (actually two measurements are made and then averaged) with the idea that things have settled down by then.

After doing a simulation since 2002, I finally "outed myself" in christmas/2004 because it was outa control - it was just a fun little hoax. While all of the above is "doable", the vast majority was also simulated. For 2005 when the webcam is real, I do display accurate weather data, but none of the other stuff listed above - maybe sometime in the future! ;-)