Timeline for What sensor to choose to track human presence indoors?
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 16, 2017 at 7:31 | comment | added | Anshul | @MaNyYaCk A very naive implementation could be simply having a bit value set to 1 when you detect motion, i.e. firing of sensors in a sequence from door to hallway (first is door sensor) - entry, and to 0 when it is the last sensor to fire. This is of course very trivial and easy to fool. (This is what is also said in the main answer by Olin above). What if more than one people entered and not all left? So, you might then want to fuse information from many sensors for a better decision model. | |
| Jan 16, 2017 at 6:33 | comment | added | MaNyYaCk | @Anshul how is that done? | |
| Sep 11, 2013 at 3:47 | comment | added | Anshul | Since you are using a microcontroller (or many microcontrollers) you might locally want to preserve the last state of the room in the control unit or the node by implementing a state machine. So, even though the sleeping person is lying still and there is a uniform (not quite) IR signature emanating off him, by knowing from the state machine that someone entered the room, the system may grow more aware of the movements. | |
| Sep 29, 2012 at 7:35 | vote | accept | Vorac | ||
| Jun 29, 2012 at 20:56 | comment | added | Dan Is Fiddling By Firelight | The non-moving person detection case can be a problem. A few weeks ago I stayed in a hotel that used motion sensors of some sort to activate the bathroom lights. About 10% of the time when I was sitting/standing motionless the lights turned off. Since they normally stayed on I assume that whoever sold them the system added logic to detect the case but it was slightly buggy. I was never able to discern a pattern in how it failed. | |
| Jun 29, 2012 at 17:00 | history | answered | Olin Lathrop | CC BY-SA 3.0 |