The Homey Community has been moved to https://community.athom.com.
This forum is now read-only for archive purposes.
The Bathroom Light Challenge
Some basic things are just difficult to fix...
I think most readers in this forum - as on other home automation fora - will know exactly what the title of this topic refers to. And, yes, you're right:
How to automate the bathroom lights _without_ the issue of "lights turning off, when things on the toilet take a bit longer than usual".
Of course, you can set the timing to "extremely long" - but that's no fun...
So, how have you done it?
Which combinations of devices and flows have you implemented to avoid forcing people to jump around on the toilet seat when the lights went off? Not using a timer setting of 90 minutes or something like that...
Searching the web, I found that the idea I got and thought was very original, wasn't exactly unheard off - but still not mainstream: The combination: force sensistive resistor + a binary sensor of some time + some modding of the toilet seat...
The challenge of course is hygiene and design - how to get it into a toiletseat, still working, hidden away and without too many wires etc...
Until I've figured this out, I'm just interested in other setups - how have you solved it?
Comments
How is this solution connected to Homey? Z-wave or?
Beside this solution, I have heard that there are sensors reacting on body temperature instead of movement (PIR). http://uk.farnell.com/omron-d6t-mems-thermal-ir-sensor
Untouched the resistance is towards infinite, pressure will decrease it to around 0 - effectively making a short circuit - which trigger the contact if connected to "Gnd" and "In"...
https://www.conrad.nl/nl/alarmmatten-750188.html
Wasn't satisfied with having to run some wire on the surface of the toilet seat, neither cutting a slit to hide the wire in and try to seal it afterwards seems easy.
Gave Junior a kit from Circuitscribe for Christmas (that he didn't like) and I took the ball pen from the set and tried to make the circuit on the toilet seat using that:
Although possible to "write" on the toilet seat, it seems like the ink doesn't "connect" and the circuit can't be made :-(
Next idea:
Got the idea that maybe coppertape could can be used and found some only six mm wide from Chibitronics as well as som conductive paint from Bare Conductive...
Proof of concept:
The two legs of the resistor are fastened only by running some tape over them, the two wires are fastened by dropping some paint over the wire while held on the face of the tape (and holding it still until the paint is dry, of course)...
Pressing the resistor triggers the alarm in Homey right away...
Next is to hide the resistor properly and get the circuit around on the toilet seat - and get it all hidden...
To be continued...
I am watching your progress as I look for such solution. Just please don't post photos with the real test!!
I have dry walls and ceilings in my house and I am thinking to have an ultrasonic sensor above (or beside) the toilet tuned to switch when there is a person on the toilet seat. Then wired into a Fibaro Universal Sensor. Problems I have is to first to find a cheap ultrasonic with white head to match the ceiling colour and second be able to switch to positive so I can use it withe the universal sensor. No contact solution, too.
And yes, a standard IR-beam broken by somebody sitting the right place would do it - but for one it is not feasible in my case due to placement of the toilet in this case. Alternatively, a ultrasound probe could be an idea, but as you wrote, neither this is "just done"...
Looking forward to hear from you with the final solution.
Sinds (most of us I assume ) put the lid up when were using the toilet and put it down when were done this would be a workable solution. Finally I would still add a timer for the times that someone (for instance a visitor) leaves the seat up for say 30 minutes to turn of the light.
An additional benefit is that you will please the misses sinds you will have to put the lid down every time for the contact to work. That softens the blow when your tinkering in the bathroom will the misses needs to use the toilet
So obvious...
That is definitely worth looking into - it must be possible to find a small enough switch that can be hidden...
Or, continuing on the thought about detecting position of the lid:
A tilt-sensor like https://www.zwaveproducts.com/shop/brands/ecolink/tilt-zwave2-5-eco-z-wave-plus-garage-door-tilt-sensor-white
This sensor, however, is rather large - anyone who knows smaller ones?
Perhaps the simplest solution is to mount a Sensative Strip (https://www.stripsbysensative.com/guard) with the large part on the lid and the small magnet on the seat - although not really hidden...
But as a cliffhanger: My toilet seat now officially "Talks with Homey" ;-)
simplified:
1. if door opened and no motion detected then start timer and turn on lights
2. if motion detected and timer running then someone in bathroom and stop timer
3. if door opened and someone in bathroom then someone leaves bathroom and start timer
For the automation of the lights I use CountDown. On motion seen by the sensor hanging in the room, and the switch was not toggled last, it starts an initial countdown for 15 seconds. On countdown start I turn the light on if it isn't yet. Once the countdown reaches zero, I add another 15 seconds if the lastMotionSeen variable is still that same room. If it isn't, it will turn off the light (if it isn't turned off yet or if the countdown wasn't cancelled by an override in the meantime).
Works a charm - except with multiple people walking around in your house triggering all kinds of things. Pretty funny to see their heads explode when I explain how this works in Flows. :-)
Could you share some examples of the flows you've made as inspiration?
Everything worked...
But:
The single connection that I knew was the vulnerable point proved to be just that :-(
The toilet seat is a "soft close" model which means the connection between the hinges and the seat and the lid are made by some ceramic cylinders with some sort of mechanics that provide resistance when stuff is rotating.
Therefore it is necessary to make a connection between to parts rotating around each other - but without increasing the friction between them...
One side is not a problem as the movement is between the lid and the hinge, but the other is between the seat and the hinge.
Made a solution with a piece of wire but it proved to be too vulnerable - and plastering parts with copper tape made friction too big and the seat for stuck halfway when it should close by itself.
Any suggestions how to get a connection between two rotating parts?
http://www.slipring.com/miniature-slip-rings.html
Could you please post an example of these flows and variables. I understand how you use these variables, am keen to try it myself, but don't know how to implement it.
Thanks!