Arlo Pro motion detection

[7gk]

I have 4 original Arlo Pro cameras. I didn't need to purchase subscription at that time, so I still haven't.  I noticed within the last year-the cameras do not pick everything up.  We have lots of wildlife 24 hours and there is always something going up.  I don't understand why it will pick up a moth flying at night but not a car or people/animals walking by.  I suspect because I don't have a subscription, that the ARLO company is not allowing all activity.  Has anyone else had this problem?

[TomMac]

I'm still using Pro cameras too.

There are many reasons for not picking up motion.

Check these things;

1) Sense level.... default is to low, IIRC it is set at 80%, many of mine need to be higher.

The sense level needs to be higher for small animals and more on their level.

2) Also the max range for people is about 20-25 ft ( better picking up if moving across the view ) read about proper install

As to the moth, yes sometimes very close insects will set off motion. PIR detection can be hit/miss at times under close really close conditions..have seen spiders walking setting mine off at times.

3) time of day ; ie sunlight shining into camera or shadows with wind

It's not being restricted.

[ShayneS]

Hi @7gk 

Have you adjusted any motion sensitivity settings or camera placement by chance? PIR (Passive Infrared) sensors in Arlo cameras detect motion by sensing changes in infrared radiation emitted by warm objects, such as humans and animals. They are passive, meaning they do not emit energy, and are optimized to reduce false alerts from shadows, lighting changes, or small objects.

While insects generally do not trigger PIR sensors due to their small size and low heat, a nearby insect could occasionally cause a trigger, especially if it is closer to the sensor than other objects in the environment. Arlo cameras combine PIR detection with pixel-based motion analysis, which monitors changes between video frames, to further improve accuracy and minimize false alerts from small or distant movements.

[Edinburgh_lad1]

So, "insects are passive and do not emit energy", but occasionally they'll trigger the sensor. Essentially, by divine providence. So,  in other words: they don't trigger sensors but they do. 

I suspect because there might have been further reduction in footage quality when recorded (past year or so), the pixel-based analysis is no longer accurate enough. 

Bottom line is: there are more reports on this forum and elsewhere that motion detection has deteriorated.

Can you confirm @ShayneS that subscription status doesn't affect motion detection please? I couldn't find anything about that in KB articles.

[StephenB]

---

@Edinburgh_lad1 wrote:

So, "insects are passive and do not emit energy", but occasionally they'll trigger the sensor. Essentially, by divine providence.

Or physics. The sensor is differential. There are two slots that are each generating voltage that depends on the amount of IR light that is reaching that slot.  The sensor subtracts the voltage from two slots, and triggers when that difference changes.

One important factor is how close they are to the sensor.  For instance, an insect could actually land on one of the slots.  In that situation, it could either block some IR from the environment (reducing the voltage), or be slightly warmer than the environment (raising the voltage).  Either way, the differential signal will change as it lands.  If that change is large enough, the sensor will trigger.  The differential also will change if the insect flies just above the sensor.

This is just one of many possible scenarios.  Swirling dust can trigger the sensor, simply because the particles can block some of IR coming from the background.  So can rain or snow.

The point I am making is that being "passive" isn't enough.  Anything that isn't transparent to IR light or which can reflect IR light can sometimes trigger the sensor.  One of my own cameras is sometimes triggered by reflections in a window (I set up an activity zone to block that).

---

@Edinburgh_lad1 wrote:

the pixel-based analysis is no longer accurate enough. 

---

What pixel-based analysis?  The 4030/4030P Pro cameras don't do any (even when plugged in).  If you mean activity zones, they are a subscription feature, so don't apply here.

[Edinburgh_lad1]

>What pixel-based analysis? The >4030/4030P Pro cameras don't do >any (even when plugged in). If you >mean activity zones, they are a >subscription feature, so don't apply >here.

Ask Shayne, who said that.

[StephenB]

---

@Edinburgh_lad1 wrote:

Ask @ShayneS, who said that.

---

Well, some of the Arlo cameras will do some pixel analysis when the cameras are connected to AC power.  But not the Pro and Pro 2 cameras.

[EFreed]

I have the same problem.   Sometimes it will detect a car pulling into the driveway and sometimes it won’t. Also, up until this past spring, the cameras would “detect” motion and proceed to record for over 4 minutes.  And there was no visible motion that triggered the recording or anything moving to keep recording for close to 5 minutes. 

Very, very frustrating. 

[jguerdat]

This sounds like it's all car-related. Cars, being shiny, tend to reflect the sun's IR rather well and may or may not reflect it towards the camera (or to something else in view of the camera which then reflects to the camera). Not detecting all of them can be due to clouds, the particular model of car (ie the shape), exact direction of travel, the angle of the sun's rays, etc., etc.

As for long recordings with no obvious motion, what was in view? If it's something that can reflect the sun's IR it may be wispy clouds varying the strength of that reflection.

In any event, repositioning the camera to eliminate extraneous areas from view while still getting coverage of what you need. ALso, does this only happen during the day? That would strengthen the sun argument. Consider reducing motion sensitivity during the day while again getting the coverage you need. You can use the schedule to automate changes like this.

[StephenB]

---

@EFreed wrote:

I have the same problem.   Sometimes it will detect a car pulling into the driveway and sometimes it won’t.

---

How far away is the camera from the car?

Is the car driving directly toward the camera?  Or is the camera at an angle?

---

@EFreed wrote:

Also, up until this past spring, the cameras would “detect” motion and proceed to record for over 4 minutes.  And there was no visible motion that triggered the recording or anything moving to keep recording for close to 5 minutes. 

---

Are you using "until motion stops" in the camera rules?  That will continue recording until the PIR sensor stops triggering.  Since that is triggered by changing infrared light (and not visible light), there might not be visible motion in the field of view. 

Not sure what might have caused that sensor to trigger in your case (and of course we can't rule out a bug).