Arlo
Support

baumbachaflocka · ‎2023-12-04

The old app worked great with geofencing, it would not arm unless every device left the area. Now it always arms if just 1 person leaves. My system is configuration is the following: anyone arrives> standby, everyone leaves> arm away

They are both set to always enabled and I have made sure that both my devices are setup correctly with always allowing precise location, notifications and do not have battery saver on. Its so frustrating because the old app worked great and the new one is so bad and we were forced to switch when we added the new cameras.

More info there is 1 iphone and 1 android phone used for geofencing, I have it set to the largest area and no other automations except for the geofencing ones.

ShayneS · 2024-09-03

Hi,

Can you try utilizing this setup for your schedules with the Arrive/Leave feature and let me know the results?

1. Select Routines > Automations > Select the (+) to create the desired schedules (When created, schedules will be listed under the Schedules section)

2. Create and enable the Arrive automation under > Select Automations > Select the (+) > Add Action > Change Mode > Select Standby > Next > Add an Action > Enable/Disable an Automation > Select the desired schedule > Next > Select Enable > Next

3. Create and enable the Leave automation under > Select Automations > Select the (+) > Add Action > Change Mode > Select Arm Away > Next > Add an Action > Enable/Disable an Automation > Select the desired schedule > Next > Select Disable > Next > Save

When completed you should see your options listed under the Arrive/Leave section under Automations

Thank you

alex_loo · ‎2024-09-20

@wenklaw wrote:

BUT if I return after 2359 and before 0500 that’s where the problem is as what will happen is the mode will just change to ‘arm standby’ and will ignore the fact I have a schedule stating the mode should be in ‘arm home’ during those times. When 2359 comes around again the automations at night will reactivate and work again.

I don't see this a problem, given the way schedules are designed --- they are supposed to ONLY trigger your "arm home" at 2359 sharp, and "arm standby" at 0500 sharp. It is NOT supposed to trigger anything, DURING the period from 2359 to 0500.

Therefore, when you are "arm away" but returning home between 2359 and 0500, you would have to manually switch to "arm home" yourself.

Alex

wenklaw · ‎2024-09-20

The problem with the new method is the way it did work was far superior in every way, I cannot think of a single benefit of this new approach.
In the old app it would switch back to the schedule and custom mode via geofencing on a users return therefore eliminating the need for manual intervention by the user which is the polar opposite of the definition of ‘automations’.

alex_loo · ‎2024-09-20

@wenklaw wrote:
The problem with the new method is the way it did work was far superior in every way, I cannot think of a single benefit of this new approach.
In the old app it would switch back to the schedule and custom mode via geofencing on a users return therefore eliminating the need for manual intervention by the user which is the polar opposite of the definition of ‘automations’.

I heard about the "really good" old app but has never tried it. When I started with ARLO back a few months ago, I was already with this "new but bad" app.

So with the old app, are you able to set up a single period (say 2359 thru 0500) and associate it to "arm home"? Note that this is something IMPOSSIBLE with the current (new) app.

So if the old app CAN do this, apparently ARLO has decided for a totally different design (polling and action during a period, vs triggering off mode change at a particular point of time). Whatever reason, this has proved to be a wrong design change.

Alex

StephenB · ‎2024-09-20

@alex_loo wrote:

@wenklaw wrote:

BUT if I return after 2359 and before 0500 that’s where the problem is as what will happen is the mode will just change to ‘arm standby’ and will ignore the fact I have a schedule stating the mode should be in ‘arm home’ during those times. When 2359 comes around again the automations at night will reactivate and work again.

I don't see this a problem, given the way schedules are designed

Therefore, when you are "arm away" but returning home between 2359 and 0500, you would have to manually switch to "arm home" yourself.

Personally I think it is a problem with the design, and that it does need to be corrected.

alex_loo · ‎2024-09-20

@wenklaw wrote:
The problem with the new method is the way it did work was far superior in every way, I cannot think of a single benefit of this new approach.
In the old app it would switch back to the schedule and custom mode via geofencing on a users return therefore eliminating the need for manual intervention by the user which is the polar opposite of the definition of ‘automations’.

just curious. with the old app, if I have been "ARM AWAY" for days and then returning home in the midnight between 2359 and 0500, I understand that ARLO will still nicely switch to the desired "ARM HOME".

Now, if I manually switch it from "ARM HOME" to "STANDBY" (for some adhoc reason), still during this 2359-0500 period, will it stubbornly jump back to "ARM HOME"?

Alex

alex_loo · ‎2024-09-20

@StephenB wrote:
Personally I think it is a problem with the design, and that it does need to be corrected.

ok now we are discussing the "better" technical design for ARLO! 😉

I started my ARLO experience only with new app. And I understand that schedules and automations are all event-driven (by clock time, geofence trigger --- respectively). It makes its own sense and I see:

PRO:

- simple, only perform actions like switching modes at certain particular INSTANCE (ie when 2359 sharp, or when iPhone GPS crosses the circle line from outside to inside)

CON:

- won't do anything during that period.

- if you do not trigger-start a period from the beginning (say returning home between 2359 and 0500), you will have to manually change to your desired mode

On the other hand, I am hearing the "old app" does not have CON above, but then I am suspecting it has its own CON --- that you can't override that "2359-0500" mode with a different mode, even if there is a strong adhoc reason.

Thoughts? Apparently I am just making some wild guess about the old app hehe 😅

alex

StephenB · ‎2024-09-20

@alex_loo wrote:

just curious. with the old app, if I have been "ARM AWAY" for days and then returning home in the midnight between 2359 and 0500, I understand that ARLO will still nicely switch to the desired "ARM HOME".

The old app interface had an explicit schedule (saying what cameras are armed at any point in time). With geofencing, you could set the "home" option to "Schedule", and when you arrived home the system would arm the cameras per that schedule.

@alex_loo wrote:

Now, if I manually switch it from "ARM HOME" to "STANDBY" (for some adhoc reason), still during this 2359-0500 period, will it stubbornly jump back to "ARM HOME"?

No, it won't. Nothing happens until the triggering time at 0500. This is something easily checked (and I have done that).

While @Jeffi has said that his system appears to be following the schedule in your scenario, that is not something I've been able to check myself, since my app is not yet properly detecting when I arrive and leave.

wenklaw · ‎2024-09-20

You could override the old app, disarm it permanently, arm it etc but you would then need to change it manually back to geofencing mode BUT 99.999% of the time you could just leave it alone, effectively fully automated.

The only ‘plus’ I see with the new app is if you manually change the mode it will then revert back to the set automations automatically next time they trigger BUT leaving the system disarmed in certain circumstances when you’ve told it you want it in a certain mode between certain times is far from an automation in my book

alex_loo · ‎2024-09-21

@StephenB wrote:
The old app interface had an explicit schedule (saying what cameras are armed at any point in time). With geofencing, you could set the "home" option to "Schedule", and when you arrived home the system would arm the cameras per that schedule.

"any point in time" means the old app needs to revise the camera status EVERY MINUTE (the technical term for this is polling every minute). Whether you like it or not, ARLO has decided NOT go for such design in the new app. With the event-driven type of design with new app, it only takes actions on time/geo triggers.

@alex_loo wrote:
Now, if I manually switch it from "ARM HOME" to "STANDBY" (for some adhoc reason), still during this 2359-0500 period, will it stubbornly jump back to "ARM HOME"?

No, it won't. Nothing happens until the triggering time at 0500. This is something easily checked (and I have done that).

Interesting. Now, given that "it won't" after such manual override, and then you leave home and return home before 0500 (I know this is should be rare for you), how will it switch?

Alex

StephenB · ‎2024-09-21

@alex_loo wrote:

@StephenB wrote:

The old app interface had an explicit schedule (saying what cameras are armed at any point in time). With geofencing, you could set the "home" option to "Schedule", and when you arrived home the system would arm the cameras per that schedule.

"any point in time" means the old app needs to revise the camera status EVERY MINUTE (the technical term for this is polling every minute). Whether you like it or not, ARLO has decided NOT go for such design in the new app. With the event-driven type of design with new app, it only takes actions on time/geo triggers.

No, it didn't have to do that. It only needed to

create hidden schedule triggers from the stored schedule that acted identically to what the new app does
look at the stored schedule when you returned home and apply the appropriate camera mode for the arrival time.

FWIW, I'm not saying the old app was perfect (and I don't think the new one is as bad as many here do).

But it is clear that using event triggers for both geofencing and scheduling is too hard. Plus it's not clear to me that @Jeffi's setup completely works. Once I get past my issue with geofencing triggers not working, I intend to replicate it, so I can try out what he's done, and maybe also set up some variations.

One option that I think would fix the design is for "Arrive" to have an option to set the mode per the schedule instead of setting a fixed mode.

@alex_loo wrote:
Interesting. Now, given that "it won't" after such manual override, and then you leave home and return home before 0500 (I know this is should be rare for you), how will it switch?<

The way I understand the current design, it won't switch until the next schedule trigger. Which matches your experience, if I understand your posts correctly.

@Jeffi's experience suggests otherwise - and if indeed it completely works there is something about the design I don't understand.

StephenB · ‎2024-09-21

@alex_loo wrote:
"any point in time" means the old app needs to revise the camera status EVERY MINUTE (the technical term for this is polling every minute).

FWIW, I have quite a few years of experience as a software developer, and understand the limitations of polling.

I want to add that the app can't do etiher the schedule or geofencing processing.

The schedule triggers need to be processed even if the phone is turned off.
Geofencing triggers require tracking state from multiple phones.

In principle, the schedule triggers could activated either in the cameras/bases or in the cloud. It'd be easiest to put it in the cloud, since I think geofencing needs to be processed there anyway. Motion events undoubtedly dominate the cloud transactions, handling scheduling and geofencing would only add a small fraction to the overall load on the cloud servers.

alex_loo · ‎2024-09-21

@StephenB wrote:
@alex_loo wrote:
"any point in time" means the old app needs to revise the camera status EVERY MINUTE (the technical term for this is polling every minute). Whether you like it or not, ARLO has decided NOT go for such design in the new app. With the event-driven type of design with new app, it only takes actions on time/geo triggers.
No, it didn't have to do that. It only needed to
create hidden schedule triggers from the stored schedule that acted identically to what the new app does
look at the stored schedule when you returned home and apply the appropriate camera mode for the arrival time.

100% agree above and while I was about to suggest an improvement to the current new app, you have mentioned that below!

@StephenB wrote:
One option that I think would fix the design is for "Arrive" to have an option to set the mode per the schedule instead of setting a fixed mode.

Exactly above! 👍

I'd say "option to set mode per clock time", to be exact.

Do you hear that, ARLO!

Alex

alex_loo · ‎2024-09-21

@StephenB wrote:
But it is clear that using event triggers for both geofencing and scheduling is too hard. Plus it's not clear to me that @Jeffi's setup completely works. Once I get past my issue with geofencing triggers not working, I intend to replicate it, so I can try out what he's done, and maybe also set up some variations.

(Great, we can discuss more technically! 😁)

I don't think using event triggers for both geofencing and scheduling too hard.

If I understand this correctly but PLEASE correct me if I am wrong --- both geofencing trigger and clock time trigger are operating system services (from Android and iOS), meaning the app simply sit there waiting for OS to trigger it, with the relevant trigger information that that app needs.

@StephenB wrote:
The way I understand the current design, it won't switch until the next schedule trigger. Which matches your experience, if I understand your posts correctly.

yes, you understand me correctly. And I don't mind such a non-perfect design (requiring manually override sometimes).

What I DO mind is --- ARLO has NOT been implementing its own new design correctly. For example, sometimes it simply doesn't know all devices have returned home. As a result, the relevant geofencing automation has NOT been kicked started.

In short, the issue is with bugs, not the design itself.

Alex

alex_loo · ‎2024-09-21

@StephenB wrote:
@alex_loo wrote:
"any point in time" means the old app needs to revise the camera status EVERY MINUTE (the technical term for this is polling every minute).
FWIW, I have quite a few years of experience as a software developer, and understand the limitations of polling.

I want to add that the app can't do etiher the schedule or geofencing processing.
The schedule triggers need to be processed even if the phone is turned off.
Geofencing triggers require tracking state from multiple phones.
In principle, the schedule triggers could activated either in the cameras/bases or in the cloud. It'd be easiest to put it in the cloud, since I think geofencing needs to be processed there anyway.

Haven't validated this, but I have been thinking schedules and (geo) automations should be done on the cloud:

logics/processing should NOT be relying on individual phone (say the owner user phone), since that phone may be offline/turned off. Meaning processing should either be done locally (within the HUB), or from the cloud. Apparently ARLO has chosen to use cloud.
Another reason is your point regarding geofencing process. Since geofencing status of multiple phones need to be consolidated for decision, it doesn't make sense to do this on a particular phone. Consolidating this in cloud makes perfect sense.

@StephenB wrote:
Motion events undoubtedly dominate the cloud transactions, handling scheduling and geofencing would only add a small fraction to the overall load on the cloud servers.

I'd say YES and NO. Motion events have nothing to do with state change/ logic processing. The captured video can be saved locally (HUB!, zero transaction traffic to cloud), or remotely to cloud (transaction traffice more demanding). The reporting of events to cloud (for further disemination to all user devices) are internet traffic un-demanding. I think this is easiest approach, and what ARLO is doing at the moment.

Alex

StephenB · ‎2024-09-22

@alex_loo wrote:
Motion events have nothing to do with state change/ logic processing. The captured video can be saved locally (HUB!, zero transaction traffic to cloud), or remotely to cloud (transaction traffice more demanding). The reporting of events to cloud (for further disemination to all user devices) are internet traffic un-demanding. I think this is easiest approach, and what ARLO is doing at the moment.

I was mostly pointing out that the cloud will get a lot more motion events than scheduling or geofencing triggers. There is some state lookup for these, since the cloud needs to process activity zones if they are set, and notifications also depend on the camera rules and smart notification settings. Of course the heavy lift there is the AI compute.

As far as local storage goes, I believe that is done in parallel. FWIW, since the newest cameras don't require bases, I think the percentage of users using them is decreasing. Either way, the camera/base do need to know whether they should stream to the cloud or not (since I'm sure Arlo doesn't want to pay for the bandwidth for systems that don't have a subscription).

We agree that the schedule processing could be done in the cameras, The camera clearly does know whether it is armed or not, since the power drain is much less when it is disarmed, and as I say above, it almost certainly knows whether it should stream video or not when it detects motion. Arlo hasn't given out much info on what is done where, so it's speculation either way.

alex_loo · ‎2024-09-22

@StephenB , given ARLO is helpless/useless if not simply ignoring this forum, I appreciate that this is the best conversation I have received from this forum so far. Thanks mate!

Alex

alex_loo · ‎2024-09-23

@StephenB wrote:
I was mostly pointing out that the cloud will get a lot more motion events than scheduling or geofencing triggers. There is some state lookup for these, since the cloud needs to process activity zones if they are set, and notifications also depend on the camera rules and smart notification settings. Of course the heavy lift there is the AI compute.

oh, absolutely --- the higher cloud traffic is there anyhow, once you are on subscription (since captures are all kept on cloud, for AI processing as well as easier playback for devices not at home)

Just curious, when on subscription, are notifications depending on the camera rules? I thought the cameras just blindly send the captures to cloud for processing, and notification after that.

@StephenB wrote:
FWIW, since the newest cameras don't require bases, I think the percentage of users using them is decreasing. Either way, the camera/base do need to know whether they should stream to the cloud or not (since I'm sure Arlo doesn't want to pay for the bandwidth for systems that don't have a subscription).

That's new to me regarding newest cameras not requiring base. Do they mandate subscription? Do they keep storage of captures?

I have been thinking the cameras just blindly send out captures for storage (to hub or, if on subscription, cloud) and/or processing (cloud on subscription only). Meaning that for users without subscription, ARLO doesn't need to pay for the bandwidth of captures, which are stored locally, in hub.

@StephenB wrote:
We agree that the schedule processing could be done in the cameras, The camera clearly does know whether it is armed or not, since the power drain is much less when it is disarmed, and as I say above, it almost certainly knows whether it should stream video or not when it detects motion. Arlo hasn't given out much info on what is done where, so it's speculation either way.

agree processing could be done in cameras, but I am feeling ARLO chose to make them simpler/cheaper by NOT doing so. Yes, they can hold simple status/logic like armed/disarmed --- and blindly capture+send, or not.

Well, apparently above is just wild speculation, based on my understanding so far. 😅

StephenB · ‎2024-09-23

@alex_loo wrote:

That's new to me regarding newest cameras not requiring base. Do they mandate subscription? Do they keep storage of captures?

I have been thinking the cameras just blindly send out captures for storage (to hub or, if on subscription, cloud) and/or processing (cloud on subscription only). Meaning that for users without subscription, ARLO doesn't need to pay for the bandwidth of captures, which are stored locally, in hub.

Bases are not needed for any the Essential models (both original and Gen 2), Pro 3 Floodlight, Pro 4 and Pro 5s. All the currently sold cameras require a subscription to get any cloud storage. Without a subscription, the cloud will send notifications, but never gets the video . In addition to avoiding unnecessary bandwidth charges, this also helps protect privacy (some users are more concerned about maintaining privacy than they are about the subscription cost).

Live streaming the camera does need to be forwarded through the cloud when you are away from home. So Arlo pays for that bandwidth whether you have a subscription or not.

@alex_loo wrote:
could be done in cameras, but I am feeling ARLO chose to make them simpler/cheaper by NOT doing so. Yes, they can hold simple status/logic like armed/disarmed --- and blindly capture+send, or not.Well, apparently above is just wild speculation, based on my understanding so far.

Honestly, I'd vote for simpler if I were their shoes. Not 100% sure if that is cheaper, but it would be easier to make reliable.

alex_loo · ‎2024-09-23

@StephenB wrote:
@alex_loo wrote:
That's new to me regarding newest cameras not requiring base. Do they mandate subscription? Do they keep storage of captures?
Bases are not needed for any the Essential models (both original and Gen 2), Pro 3 Floodlight, Pro 4 and Pro 5s. All the currently sold cameras require a subscription to get any cloud storage. Without a subscription, the cloud will send notifications, but never gets the video . In addition to avoiding unnecessary bandwidth charges, this also helps protect privacy (some users are more concerned about maintaining privacy than they are about the subscription cost).

Live streaming the camera does need to be forwarded through the cloud when you are away from home. So Arlo pays for that bandwidth whether you have a subscription or not.

So for these latest Essential models,

Base is not needed but --- is subscription a MUST then?
MUST I go for subscription, if I want to see captured videos?

@StephenB wrote:
Without a subscription, the cloud will send notifications, but never gets the video . In addition to avoiding unnecessary bandwidth charges, this also helps protect privacy (some users are more concerned about maintaining privacy than they are about the subscription cost).

re-asking the same question: without subscription, are you able to watch the captured videos?

@StephenB wrote:
Live streaming the camera does need to be forwarded through the cloud when you are away from home. So Arlo pays for that bandwidth whether you have a subscription or not.

That is a little design surprise to me. Why can't this streaming direct from the cameras? Just like playing back the local storage from HUB.

Alex

StephenB · ‎2024-09-24

@alex_loo wrote:
That is a little design surprise to me. Why can't this streaming direct from the cameras? Just like playing back the local storage from HUB.

Playing back local storage when away from home requires either port forwarding or a VPN. That is because the camera is behind a NAT (and often the phone is behind a different NAT). That approach fails if the ISP is using CGNAT (carrier grade NAT) or equivalent. For instance, neither port forwarding nor a router-based VPN service can be used with T-mobile Broadband, since there is no way to make a direct connection to your home router over that network.

So the only way to make sure that you can make the connection is to deploy a transversal server that forwards the traffic through the cloud. Often that is only used when a direct connection cannot be made (and I think Arlo is already doing that - local livestreaming is I think done directly from the camera to the phone).

IMO Arlo should have used the same transversal server they've deployed for livestreaming to access local storage.

@alex_loo wrote:
re-asking the same question: without subscription, are you able to watch the captured videos?

Only if you have a base (and of course the camera needs to be connected through the base).

alex_loo · ‎2024-09-24

@StephenB wrote:
@alex_loo wrote:
re-asking the same question: without subscription, are you able to watch the captured videos?
Only if you have a base (and of course the camera needs to be connected through the base).

so my question was under the context of "latest ESSENTIAL cameras do not need HUB". If I buy these cameras and NOT the HUB, but then do NOT get subscription,

can I watch captured videos from home?
can I watch captured videos from outside home?

And then my next question is:

can I still buy a HUB and have the cameras connecting to it (for local storage)? (under the context of NO subscription)

Alex

StephenB · ‎2024-09-25

@alex_loo wrote:

@StephenB wrote:
@alex_loo wrote:
re-asking the same question: without subscription, are you able to watch the captured videos?

Only if you have a base (and of course the camera needs to be connected through the base).

so my question was under the context of "latest ESSENTIAL cameras do not need HUB". If I buy these cameras and NOT the HUB, but then do NOT get subscription,

can I watch captured videos from home?

can I watch captured videos from outside home?

Sorry, I thought my answer was clear. When an Arlo camera is connected to your home wifi, then there are no captured videos unless you have a subscription. So there would be no captured videos to watch.

@alex_loo wrote:
And then my next question is:

can I still buy a HUB and have the cameras connecting to it (for local storage)? (under the context of NO subscription)

The Gen 2 Essentials don't have support for smarthubs yet. Arlo has said they will add that feature, but they haven't committed to a date here. Hopefully it will be soon.

After they add that feature, then you could use local storage by pairing the camera with with a VMB4540 or VMB5000 smarthub.

alex_loo · ‎2024-09-25

@StephenB yes your answer was clear enough. It's too unbelievable that I want to confirm.

Can't imagine there is such product that says you can skip subscription and then when it reports something, you can't see what has been reported.

Alex

Ps: appreciate your patience with me mate

Jeffi · ‎2024-09-25

@wenklaw yes I think I get a similar anomaly if we all go away for more than 24 hours.

StephenB · ‎2024-09-26

@alex_loo wrote:

Can't imagine there is such product that says you can skip subscription and then when it reports something, you can't see what has been reported.

Have you seen this?

https://kb.arlo.com/000062312/What-options-will-I-have-when-the-Arlo-trial-ends

It gives more details on what you lose when you don't subscribe.

There are some situations where you don't need (or want) storage. For instance, if you use a camera in a baby's room, you usually just want the live view (and likely are turning off notifications). FWIW, even if you cannot see any captured video, you can immediately look at the live camera view when you get the notification.

That said, Arlo's business plan depends on subscription revenue, so they do what they can to get people to subscribe. While I personally think they are too restrictive on local storage, we all have a vested interest in their success. The cameras would be useless if their cloud shuts down.

Arlo Support

Arlo App for Support

Arlo
Support

Arlo App
for Support