Bed Presence for ESPHome by Elevated Sensors

The device comes with short extension leads that should enable you to mount it under the slat.

I’ll check how hot mine is when I get home tonight but I doubt it will be an issue.

Ah awesome - the extensions leads should be perfect. Thanks for replying! Cheers

Just got home. The sensor is basically at ambient temperature. Not running hot at all.

Thanks @tom_l for providing some quick info. I wanted to add some numbers to the conversation to ensure you that it’s safe to run under the mattress. The ESP32-C3 chip has a built-in temperature sensor used to measure the chip’s internal temperature. While the internal temperature of the chip will be higher than the outside of the sensor, it gives us a good idea of if the chip itself even gets very hot. I also added a temperature probe to the warmest part of the exterior of the case.

For the setup, I tried to make it worst case scenario. I set the sensor reporting such that it was calculating and sending values every second. WiFi transmitting is probably the most energy intense operation, so keeping it up and running should generate the most heat. I also sandwiched the unit between a dense blanket and a memory foam mattress, leaving it almost no ventilation.

We are planning to fully support box spring setups at some point in the future, so this is a good test to verify that it stays cool even when it doesn’t have much ventilation.

Even under those circumstances, the chip itself topped out around 100 °F (~38 °C) and the warmest part of the outside topped out around 90 °F (~32 °C). Not quite ambient temp, but not very warm either. It only draws ~80mA @ 5V meaning there’s not a lot of heat to actually dissipate.

1 Like

Awesome, thanks very much for the comprehensive reply.

Looking forward to the box spring option.

For those interested in the box spring add-on, sign up for the waitlist @ Box Spring Kit (COMING SOON!) — Elevated Sensors.

2 Likes

Received my sensor, set it up in about 5-10 minutes and installed in another 10 minutes.
Absolutely fantastic device. I am very un-DIY when it comes to somethings, being very time poor doesn’t help. This was just seamless. Highly recommend. Now, to order two more :smiley:

2 Likes

Has anyone got any handy ideas around automations for this sensor? I have it setup to
Trigger night mode and eventually lights off once our phones go on charge, but I am finding the automation fails because as we move around in bed, the trigger of us being in bed or “clear” keeps happening so it restarts the automation and once it does that, certain conditions fail because they’ve been turned on from the first time. Any ideas how to reduce triggering of the sensor?

Can you share what your Pressure graph looks like over a given night (include before/after getting in bed as well) and what your Trigger Pressure is set to? It may be that you simply need to manually lower your trigger pressure.

Also, what is your Firmware version?

Sure thing Stephen!

Is this what you’re after?


Sorry - here’s a view of the graph from 8pm last night until this morning

Cool, and what about your Trigger Pressure? That’s down in the Configuration section (a little lower down from your first screen shot).

Whoops, sorry!


Your Left Trigger Pressure seems pretty good actually, but the Right Trigger Pressure seems a little high. Try setting your Right Unoccupied Pressure at 0% and that will move the Right Trigger Pressure closer to the 70% mark.

Additionally, since you’re running the DIY firmware, there are a couple substitutions you can use to increase stability at the cost of reaction speed. In your config file, consider adding/modifying the following substitutions. Below are the defaults if you don’t pass any substitutions in. By default, 5 averaging_window_samples averages the sensors over 2.5 seconds, and the 1s standard_delayed_off means the sensor must register below the Trigger Pressure for 1 second before turning off. Increasing either of these will make it so that it takes longer to register “out of bed” and will increase resilience against shifting in bed causing false negatives.

substitutions:
  averaging_window_samples: '5'
  standard_delayed_off: '1s'

Thanks Stephen! I did wonder about the right side, I might calibrate that again first as you say. With those other sensor figures, should I just double them?

You can always manually adjust as well. The auto calibration just tries to get you a good starting point. If you change one of the Occupied/Unoccupied values, it’ll recalculate the Trigger Pressure as well.

Doubling might be a good starting point. You might want to zoom in on your trouble points and look at the timestamps to give you an idea of how long the sensor registers “out of bed” when you’re shifting around. This might give you a better idea of the timing you need to set.

If your desired automations don’t require any kind of speed, you could get crazy stability by increasing the delay much more.

Perfect. Thanks for being so helpful! When I get home I’ll tinker with these changes and see how it goes for a few days :slight_smile:

Hi Stephen. Hi have already registered for ordering one of the Bed Presence for ESPHome. Is there an ETA when this item will be ready stock again? :thinking:

More will be assembled and tested within the next week. Just in time for some early Black Friday sales!

2 Likes