Interference and crosstalk: Both 24GHz and 60FHz mods are tolerant of other emitters in their field of view, so you could pair two of them in the outside corner of your ell, looking towards the ends and a third at the long end of the ell looking back towards the corner and they ought not to interfere. Another possible configuration would be three ceiling mounted emitters looking down, overlapping the fields of view for full coverage of the floor area. A 60° field of view at 2.8m gives you roughly a 5m radius of detection on the floor.
Seeing through walls: “Thin walls” attenuate the signal, but allow enough penetration to detect people through cover. The radar can penetrate thin plastic housings, plexiglass, cloth, drywall, wood paneling, and thin plaster & lath. It cannot penetrate glass, stone, or metal. Same goes for floors. My radar module will detect people in the unfinished room below my home office through the hardwood floor if I angle it down far enough.
Fast detection: I found that while the radar was responsive even to small movement nearly instantly, it takes up to 2 or 3 seconds to acquire and classify a target as a person. Once t locks on, it pretty reliably tracks the person for as long as they are in view. In practice, a person walking into the edge of the field of view at a normal walking pace could cover half the field before they are detected as present. It “feels” a bit slow compared to PIR detection which is sub-second in most cases, but generates a lot of false positives. The technique I am refining is to position the the radar to detect a person entering a room by angling the radar field of view to “lead lead target” and use the PIR to determine the target has moved into a zone where I want action to take place. In essence, I want the lights to trip on when the radar detects human presence AND the PIR sees a hot blob. Then I kick on the light and wait for the human presence detection to go back to unoccupied for 5 minutes before turning off the light. I might also experiment to see if ultrasonic detection is any more reliable than IR, since it wouldn’t be fooled by warm/cold draughts, or by sudden changes in light as from dappled shade or clouds & sun. My concern is I have pets and wouldn’t want to stress them with sound I can’t hear, but maybe they can. Need to research it more.
Its been fun playing with this stuff, but I might note that at this point you can just buy an open platform (as in open source) esphome/HA compatible multisensor presence detector made by the guy who does the Everything Smart Home channel on YouTube. He posted a couple videos talking about it (links below). For me, this is just messing around with something I always wanted to play with, but I’ll probably just buy a kit when I want to hang something functional on a ceiling or wall.
Definitely check out the videos I linked. They’re excellent for helping to understand the pros and cons of this tech in real-world smart home applications.
Everything Smart Home - Building my own smart ho e oresense sensor
I looked further into my assertion that 24GHz radar was being phased out, since something was bothering me about it. It seems 24 GHz Ultra Wide Band (UWB) applications are no longer approved, but 24GHz Industrial, Scientific & Medical (ISM) applications are still approved. Home use of 24GHz radar would be considered ISM, so there is probably no restriction on its use.
I’ve been messing around with the MR60FDA1 60GHz mmWave detector in esphome and Home Assistant. Unfortunately it has the same 6m range and 60° field of view you’re looking to get around.
I will say, though, that within range the sensor is quite responsive, and detects static presents quite well. The high frequency gives it sufficient resolution to detect micro movements like fidgeting, looking around or even breathing. My module has fall detection on board, purportedly to sense if a detected person is standing or laying down. Another version supposedly can detect respiration and heartbeet within a couple of meters.
The good thing about 60Ghz is they tend not to interfere with each other, so several units could be arranged with overlapping fields of view.
I’m pairing mine with a PIR module for rapid detection, and to help eliminate false positives on radar hits since radar can see through walls, and doesn’t necessarily expose the distance to the target in esphome.
The 24GHz models have a longer range up to 12m and may have 360° fields of view, but have lower resolution and ranging for micro motions. In the US, they are being phased out for potential interference with aviation though I can’t speak for other countries.
Andreas Speiss posted a good video on YouTube that covered a bunch of different models (link below) that I thought was informative. It will.lead you to some other similar content that might help you to assess your needs and match a product to your application.
You might be able to do something with local Zigbee2mqtt or Zwave2mqtt proxies at the cabin, and tie those areas back into the main instance over Wirguard or Zerotier, but you’d need a pretty reliable internet service at the cabin to do so.
I don’t use those proxies myself since ai just use the one local hub, bit it should be possible. Maybe someone with more experience with those proxies could chime in?
Oh, I never said they weren’t absolute prats about invading user space with advertising their bullshit. The Lens fiasco, Snaps, the popup warnings in apt breaking scripts, and the lack of UI toggles to easily disable those nag messages are all reasons I run other distros. There’s a big Mint colored button to turn on the Ubuntu experience without the nagging.
You have other choices that do no not shove that bullshit in your face. Canonical is gonna canonical. Nobody said you have to play their game.
My point was they are not withholding anything community-based from anyone. They are entitled to charge for their original work, even they are pushy about it. They even abide by the license and distribute it the changes when complete, but they’re not gonna just do it for giggles.
Not that I’m a fan of Ubuntu here (I generally don’t run it when I can run anything else), but I do want to say I think you’ve missed the point of the Pro tier.
Ubuntu releases two stable versions a year which are supported for 2 years or so. This is like a slow rolling distribution, and makes the newest software’s available. It receives regular security updates from upstream, from Canonical, and from backports, again for up to about 2 years. Most users install this version.
Ubuntu LTS editions are similar to the above, but receive all the same security updates for 5 years instead of 2. These distributions are generally targeted for Enterprise users who value stability over having the newest software, and for whom upgrading comes with significant time, expense and risk. The 5 year window is customary among other distros, and is largely supported by and throughout the Dev community.
Ubuntu LTS Pro editions extend the LTS support editions for an additional 5 years, meaning a Pro distro enjoys 10 years of security updates from upstream, backports, and from Canonical where needed. Canonical might even open source their fixes back into upstream for other maintainers and distros to use, depending on the situation. However, since Canonical is providing the work, they charge subscription fees to cover their costs for it from their target audience: Enterprises who can’t or REALLY don’t want to upgrade
Why an Enterprise might not want to upgrade has to do with risk and compliance. Corporate IT security is a different world, where every bit of software may need to be reviewed, assessed, tested and signed-off upon. Major software upgrades would need to be recertified to mitigate risk and ensure compliance, which takes significant time and expense to complete in good faith. Not having to do it every 2 or 5 years is money in the bank, especially when the environment doesn’t introduce new requirements very often.
Canonical is meeting a market demand with their Pro tier by allowing these customers to spend a fraction of their recertification costs on a software subscription. It’s overall good for the ecosystem because you have what amounts to corporate sponsors pumping money into keeping older packages maintained for longer. This let’s them keep using the same software distro all the rest of us can use for free.
I’m not shy about calling bullshit on ANY distro that operates in bad faith, and they all get into some BS from time to time. Nevertheless, Canonical are acting in good faith on this, and are merely collecting money for their time and skill to provide maintenance on FOSS packages that might otherwise go unmaintained.
tl;dr: Pro tier is for Enterprise customers who need extra-long term support and are willing to pay for it. Canonical is meeting a market demand so they can remain competitive for use in those environments, which is good for everyone. It’s benign. Keep the pitchforks sharp and the torches dry for another day.
edit: The following is off topic, but I’ll.leave it as a testament to my gray-beardedness. In my defense: Unity isn’t Unity anymore. Don’t get old.
I’ve been using Linux for 30 years now, and for a while I was an advocate for Ubuntu and Canonical (among others, I’m pan-distributive). Then things changed: GNOME 3, Wayland, Unity, something-sonething, Snaps… All too much.
As an advocate, I’m apt not to emerge with favorites, or to yuck others’ yums. Neverthekess, Canonical is a press beyond the pale, many days.
In the end, I don’t recommend Canonical distros. LMDE is solid, as are most of the *bian and redhat downstreams. I don’t recommend the others because I don’t know them, but more importantly I couldn’t help a friend un-bodge a bad installer on them (likewise for "BSD or Darwin).
But really, no love for Canonical. They went to some Dark Side, and I’ll have a hard time forgiving them for it.
I pay for Hulu and Prime (for the shipping), and get Netflix & Apple TV thru my T-Mobile subscription.
edit: oh, derp. I also have YouTube premium because I paid up for the Google Play Music family plan back in the day, and then it converted to YouTube Music + Premium. Later, I upped my subscription to Google One for the extra storage and phone support and get some okay, ad-free free content on the 'Tubes
I think the i5 is Ivy Bridge, but I couldn’t tell you what gen that is. My main use of HA aside from the automation is Frigate, which apparently needs the hardware AVX flags. This chip supports AVX512, where my older AMD did not, so that’s why I went with it. Its an i5-3470T, if that helps.
I’ve got HA with Frigate + USB Coral w/4 cams, FlightRadar24 receiver/feeder, ESPHome, NodeRed, InfluxDB, Mosquitto, and Zwave-JS on a refurbished Lenovo ThinkCenter M92p Tiny, rigged with an i5 3.6GHz, 8GB RAM and 500GB spindle drive. It’s almost overkill.
Frigate monitors 2 RTSP and 2 MJPEG cams (sometimes up to 3 RTSP and 5 MJPEG, depending of if I’m away for the weekend) with hardware video conversion. FR24 monitors a USB SDR dongle tracking several hundred aircraft per hour. I live under one.of the main approaches to a major US hub.
Processor sits at 10% or less most of the time, and really only spikes when I compile new binaries for the ESP32 widgets I have around the house. It uses virtually none of the available disk. It’s an awesome platform for HA for the price.
I’ve been using the Sendled Sengled E11-N1EA bulbs available on Amazon for quite some time now. They have a bit of a cold white edge to them and could maybe be a bit brighter at times, but overall I’ve been happy with them.
Mine came in a 4-pack. I have two inside in table lamps, and two outside in coach lamps.