Home/Away status - Fine tuning #595
Replies: 5 comments
-
Hi Paul! The first thing to keep in mind is that the only "fact" that bermuda gets from the world is "I just saw this thing". Nothing ever tells us "I can't see this thing", because we have no "proof" of a thing being not there. So all we know is that at times, we know the thing is there, but at other times we get nothing, and at some point we need to decide if we think it's gone. For the You can reduce that timeout to whatever value suits your particular use-case. As with everything in life though, this is a trade-off. The shorter you set the timeout, the more likely (and more often) you will get "nuisance triggers" - bluetooth is a "best effort" thing - devices just spit out advertisements, and maybe we get them, maybe we don't. So you can adjust the value down until you start getting false indicators for If you want something that will lock doors for you etc, then you might be better off using a trigger that is conclusively affirmative (like "the lock button was pressed") rather than an assumption ("haven't seen the car lately..."), but it all depends on what works for you. |
Beta Was this translation helpful? Give feedback.
-
Hi Ashley, |
Beta Was this translation helpful? Give feedback.
-
Latest update. |
Beta Was this translation helpful? Give feedback.
-
Reading this information from you, even if I'm a few months old, I understand that it's normal to have to increase the consideration of seconds to consider myself "out" in 170s, right?. I understand that these "problems" happen because there is also a clock connected to bluetooth, and when the data is sent from the mobile sometimes the proxy may not listen to what the phone sends And I ended up thinking that it is outside, however it seems tremendously broad to me for all that period that an active listening of something that I imagine takes time cannot occur, but also Must be affected by the Bermuda listening time value, which I had set to 1 second, but I have increased to 5 and may increase to 10 to continue the tests properly. Thanks for reading! |
Beta Was this translation helpful? Give feedback.
-
Hi Paul, not sure if you're still battling this, but another option to consider is to see if you can place any proxies at the end of the driveway near the street - like in a letterbox. If that's possible, you could:
You'll still need to think carefully about different corner cases, like if you just drive past the house and continue on, or if say an upstairs bedroom proxy can "see" your car driving down a street that backs onto the property etc. Hopefully that's helpful. Since I've enabled the discussions feature on the repo, I'll convert this to a discussion topic rather than an issue, so please feel free to continue the chat over there 👍🏼 |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
This is my first contribution. Sorry if it's not exactly set out as it should be.
I have Bermuda running on an HA Blue (yes they still exist!). I am using Blue Charm ble beacons as the source, fitted to cars in my driveway. I am only using detection for the cars being on the driveway or not on the driveway.
With my current settings Bermuda reports devices as being 'Home' (when they are home) consistently (so far).
Bermuda also reports devices as 'Away' when the cars leave the driveway and 'Home' when they return. So far, so good.
When a car returns to the driveway, Bermuda picks it up quickly and correctly reports 'Home'.
When a car leaves the driveway, however, Bermuda is slow to detect it leaving and hence slow to report it 'Away'.
As I'm not sure exactly how Bermuda evaluates the data to report Home/Away, I'm not sure what adjustments should be made to improve the 'Away' detection.
Can you point me to any information on Bermuda's calculations for the 'device tracker' which will help fine-tune it's responses?
Beta Was this translation helpful? Give feedback.
All reactions