No-one is blaming the Wyze camera. It doesn’t matter what the cause is, a drop out is a drop out and if the camera fails to reconnect, then you have a down camera that’s not recording anything. WiFi interference is common and is nothing to do with the camera. If you look at your internet logs, you’d be surprised how many times your router may experience connection issues of one type or another. Ignore the up time log and look in the detailed log. You’ll probably find many drops in any 24hr period even if only for a few secs.
Even ignoring reliability, as said before, wifi has limited bandwidth once you start spreading it over several cameras. 2.4ghz even with one device can go as low as 15mbs. Just try monitoring your wifi connection to eg a pc with software in real time, it’s an eye opener. You can forget seeing 433mbs. That’s theoretical. You might see around 1/3 rd of that at times. However, around 1/10th of that isn’t uncommon. It’s widely reported by many users and the further you go from the router, the slower the connection. Here users are reporting drops as low as 15mbs:
I’ve been on my laptop with quality USB WiFi adaptor, external aerial and been within 15 feet of the router and have seen speeds as low as 45mbs. That’s about enough to support 3 cameras without anything else using the wifi or the wifif speed dropping any lower. Personally, I’m not against wifi, it’s just limiting and if it’s the sole connection means, it’s a potential reliability and speed bottleneck.
The issue with solar is cost. It’s not as simple as just a small panel.
You need a panel capable of keeping the battery charged at all times. This varies according to location and hours of sunshine. But necessitates a large panel in many locations (by large I mean not compact ie maybe a few feet square, not roof sized). In sunny areas you may have issues dumping charge. In less sunny areas, issues getting enough. This makes it very difficult to produce a one size fits all product.
Assuming it can be done, you have the panel cost plus a mounting bracket the cost of which will vary according to panel size and where it needs to be mounted. You the need a charge controller. You then need a battery of a size which ensures it never goes below 50% charge even in the coldest locations - again climate affects charge levels. You then need something to house the battery assuming it’s too large to go into the camera which it probably is by a long way as it’s probably the size and weight of a small leisure battery. Then you need suitable wiring. Solar transmission losses are high so the higher voltage, the less the power loss but that needs thicker copper cabling for the longer the run and bearing in mind you’re unlikely to mount a leisure battery up a wall due to size and weight… High voltage charge controllers tend to be larger and more expensive. The costs mount up. I believe someone above calculated a system out and it can out from memory in excess of $150 just for the solar component.
Non solar batteries do have a further issue. Even with a long life one, you need to know when it’s getting low and remember / be bothered to change it otherwise your camera will go down. Again, not against battery. I recognise battery is important for some uses Wyze cameras are used for. However, it’s not the best solution for security purposes especially for an external camera as batteries also drop off dramatically in charge when exposed to the cold (hence why most cars that fail to start, do so in winter).
The cams do the constant record no matter what. According to my router logs the entire problem has been my ISP . DDos attacks have been quite prevalent lately. Since theres no internet, you get no motion flagged segments requiring one to review hours of video from the chips. A smart NAS could change all that.
Ya, I’ve specd out cable for offices and industrial sites. I initially didn’t know the designations but like everything else that was new to me I read and learned. In the 90s our office was transitioning from an old legacy ibm system 36 with twinax connected terminals to PCs and Ethernet - we always tried to do things safely. The only thing I had to hire a contractor was for running fiber on our utility poles to connect the remote building switches. We did all the rest in house.
Solar can be a solution, but it’s not a solution everywhere and it is more expensive.
Take my north facing front porch for example. The only way that solar would work there is if the panel could be mounted remotely in which case it would have to be mounted to a eave that gets some sun and then a wire run down to the camera. Hardly ideal and you’re really to the stage at that point were you might as well run that wire to mains power or do a POE drop rather than fool around with a solar panel.
On the other hand, if I wanted a camera at the rear of my property on an unpowered outbuilding, then solar is pretty much the only game in town if I want real time monitoring. But, I’m going to pay for it.
And yes, you do have to size that panel and battery for the minimum number of daylight hours in the winter and however many cloudy days you want to be able to bridge.
Let’s say the shortest day of the year is 6 hours (could be more or less depending on where you live). I think a Wyse Cam is about 2.5W, but lets say you could get that down to 2W. You’re going to need a panel capable of about 10W minimum just to bridge the usage during the night. If you want to assume cloudy days (we do tend to have those in the winter) that could easily go up by a factor of 2 or 3, so lets say 20 - 30W. Now you also need a battery to store that energy. Maybe a 15Ahr battery of some sort. Lithium if you want something small and not overly heavy.
Now this is for a camera that gives you the same level of functionality as the current Wyse cam. On all the time, and on your wifi network all the time. You could stream continuously or initiate streams on demand. If you’re willing to accept motion event driven streams only and no streaming on demand you can get by with a lot less.
Yep apart from the battery couldn’t be 15AH, because you can never use the full amount of the battery without damaging it’s lifespan. Typically with some types, you can never discharge below 50% unless you want to severely affect lifespan. So your 15AH battery just became a 30AH battery. Then allow for cold weather in winter reducing capacity… You’re soon at a 40-50ah battery which is the size of a small car battery.
Yeah, I didn’t do the back of the napkin math on the battery so you may well be right on that.
Things add up quickly if you’ve got a device that pulls a couple of watts and might be mostly in the dark for a couple of days or three.
I’ll say this - if Wyse could figure out a way to quickly connect to a wifi network, check if there is a request to stream, then get off the network and shut down such they could do that often enough that stream requests are reasonably responsive but with a low enough duty cycle on the wifi transceiver that you don’t kill your battery then they’d have a leg up on the competition. Then you’d have motion event driven streams (via a very low power PIR sensor) plus the ability to command a stream so you can look around on demand.
I tried to design a similar product (not a camera but another IOT device) about 4 years ago and I could never make the power requirements versus the responsiveness work out. Maybe wifi transceivers are more suited for this application today, but I sort of doubt it. Wifi was never designed with low power in mind. I ended up using zigbee but you’ll never get the bandwidth to do any decent resolution video through a zigbee link.
Fwiw, for our needs I can live with PIR events only, and maybe just 30 sec of live view on demand every once in a while - especially by being able to mount in a better area so I can avoid the incessant vehicular traffic events I now get due to inability to better angle the cam inside. However I also just realized I can get power in a few locations from the flood lights wiring mounted under the eaves.
It’s the on demand that’s the problem unless Wyse has built a better mouse trap.
To make on demand work, you’ve got to have the camera’s wifi on, at least the receiver - so it can hear the request to stream. That’s a problem because even just the receiver pulls a fair amount of power compared to the battery capacity. The PIR can be made to pull microamps of current. The wifi receiver is more like 10s to 100s of milliamps. This will really drain a battery if it’s present all the time.
All the battery powered cams have a problem with this. I helped a friend install a Ring Doorbell in a location where there wasn’t an existing wired doorbell to pull power from so it was battery only. The fact that he couldn’t command the camera to stream on demand was his biggest complaint.
Maybe Wyse (or Xaomi) has come up with a way to have the camera periodically connect to the network, see if there’s a stream request, then shutdown. I’ve tried to do this on a product and I just couldn’t get the duty cycle of that low enough it would give good battery life. It’s all tied up in how quickly you can connect, check, then shutdown versus how often you can afford power wise to do that. You’d like to do it maybe every 5 or 10 seconds to be reasonably responsive. But, you need to have that receiver off 99.99% of the time. Hopefully Wyse has come up with something unique.
I suspect they will be using a gateway to connect to the camera with either an existing low power consumption protocol or their own (new) protocol. I agree that using WiFi for a battery powered camera will be extremely difficult to manage power consumption without sacrificing reliability/features.
Yeah, low power wifi wasn’t really available when I looked at this problem. Things have improved. Here’s the power numbers for a Silicon Labs low power wifi transceiver.
Power Consumption: * Standby: 3.8 uA
Sleep: 10 uA
Wi-Fi Powersave: 0.77 mA
Active Rx: 5.7 mA (1 Mbit/s UDP)
Active Tx: 11.4 mA (1 Mbit/s UDP)
There’s only two numbers that are important there - Powersave (0.77mA) and Active Rx (5.7mA). Still, that’s a lot more power than most people would believe. You want an average current on the order of microamps. Something powered off AA primary batteries aren’t going to last long with those numbers. But maybe if you put a big enough lifepo4 battery in there you might make it work. Maybe it would be ok for something on your front door. You wouldn’t want to mount something like that under your eaves though where you’d have to climb a ladder to change or recharge batteries.
The few I’ve seen online seem more like 1-3 mo life for the rechargeable battery. I guess it depends on frequency of the PIR triggering it to connect to the WiFi. Personally I’d hate to deal with a monthly battery swap or recharge unless it was a doorbell type cam that was easy to access.
This is just going to cause a bigger flaw than what the indoor cameras already have…
By the time the PIR detects motion, send the signal down the trigger while, the code interrupts that signal, starts up the IR LEDs, camera focuses, and the recording starts… the moment that was suppose to be captured is gone…
This is a similar flaw in the indoor cameras that do software based motion tracking… The recording starts seconds after the motion is first detected… So the UPS guy has already waked down the the driveway, dropped the package off, and is heading back to his truck before the recording even starts… Or the dog runs through the living room, and all you catch is a blurred glimps of the tail turning the corner down the hall…
Why don’t these cameras write a small buffer to the internal memory (while it small, there is still enough space for a 1-2 second buffer… or even better if there is a microSD use that to cache a video buffer to, and then when motion is trigger, start recording 1-2 seconds in the past…
That’s normally battery cameras in low power mode with the camera off & only the pir on, on motion the camera has to startup, if they did what you want they would be flat in a day or less
Wired systems constantly record & the pir is only to flag the timestamp as a possible interest
@WyzeDongsheng@UserCustomerGwen I’m so glad this product is in development! I own 5 Blink outdoor cameras, but I like Wyze’s interface and features so much better. I’ll keep my suggestions limited to software, since it sounds like the hardware is already pretty far along. And it sounds like the main hardware suggestions I would make are already being implemented: Battery-powered with hardwire option, and video storage on indoor hub.
From a software perspective, here are some things I hope to see. For me, 1-4 are pretty important. 5-7 are “nice-to-haves” but would really set these cameras apart,
Feature parity with existing Wyze cams, when plugged into a power source
I assume this is already the plan, but I didn’t see it spelled out in detail, Specifically I’m thinking about continuous recording options. Obviously, I know this isn’t possible while it’s operating on battery power. Which brings me to…
Option to use “battery” mode when hardwired
This might sound weird, but there’s a good reason: I already read that you guys aren’t planning to introduce a solar option for V1 of this product. But people may want to plug a separate solar panel into the hardwire connection and power it that way. If so, it would be useful to choose which mode the camera runs in. Depending on the specific solar panel and its power storage, there may not be enough power to run the camera continuously at full-steam, but if the camera runs in “battery” mode while connected to a solar panel, people could basically avoid the need to change batteries, even with a small panel.
Option for external events (including from other Wyze cams) to trigger recording.
I guess this also goes under “feature parity,” but this is one of my biggest gripes with Blink cameras. There’s no way to trigger a recording based on an outside event, aside from manually opening “Live View” in the app. You guys already have a great ecosystem set up for that at Wyze, and it seems to be implemented well on the indoor cameras. For the outdoor cameras, it will be even more important, since there can’t be a continuous recording option when they’re on battery power, and it can be tricky to perfect sensitivity levels on PIR sensors, especially if the camera faces a road. (As many of mine do, unfortunately) This will allow people to place motion sensors or other cameras in more advantageous locations, and trigger recordings to start on the battery-powered cameras when OTHER devices detect movement as well.
Event notifications/triggers happen ASAP
I wouldn’t even think to mention this, because it seems so obvious, but for whatever reason, Blink’s developers decided that their system won’t notify you of an event until AFTER it’s over. Even their IFTTT trigger which is ostensibly based on camera motion is ACTUALLY based on whenever the video gets uploaded – not when the motion happens. I guess it’s because they’re REALLY preoccupied with battery life. Since it takes a little more energy for the camera to connect you to live view, they don’t want to tell you about an event until they can show you the video of it in the cloud. While I understand the battery concerns and limitations, I think this takes it a bit too far. It’s important to know about an event WHILE it’s happening. And to the extent that certain features affect battery life, it’s better to give the user the warning and the option, rather than removing the option entirely.
Option to visualize PIR input
Speaking from experience/frustration with optimizing the motion detection on my Blink cameras, it would be priceless if it were possible to visualize what the PIR sensor is seeing. My property faces a road that gets a fair amount of traffic, and the PIR sensors give me tons of false positives from passing traffic, even when I make sure that the road is not in the camera frame. Of course, the PIR sensor’s field of view isn’t going to mirror the camera lens exactly, so it would be useful if it were possible see what the PIR sensor captures, I wouldn’t necessarily want to capture this all the time, but having the ability to enable it for troubleshooting purposes would be SUPER useful.
Smart detection to avoid false positives when headlights enter the frame
This is more about hardwired cameras, since battery-powered ones wouldn’t get triggered by this in the first place. Basically, if a camera faces an area where headlights are shining in from offscreen, pixel-based motion sensing makes it seem like a LOT of movement, when in reality, of course, there is no movement at all. The addition of the PIR sensors should make it easier to flag this as a false positive. (Basically, if X% of the pixels are changing drastically, but 0% of the PIR sensor is changing, it’s probably headlights.)
Outdoor motion sensors
Okay, this one isn’t software, but it also doesn’t affect the hardware of the camera in development. It’s a separate product. If it happens at all, I realize it would probably be further down the road, but it would be really great to complement the outdoor cameras with outdoor motion sensors. In the meantime, maybe I’ll just try to seal up the indoor ones with silicone. Haha. As I mentioned, my property faces a road that gets a fair amount of traffic, so it would be useful to be able to place motion sensors more strategically. I could then place a motion sensor by the entrance of the driveway facing the house, and it could trigger the cameras which are ON the house, facing the road. That would let me turn the sensitivity DOWN for the on-camera PIRs, but turn it UP for the motion sensor which is less likely to get false positives.
–
Overall, I’m really excited to see the product and will gladly ditch my Blink cameras as soon as this is available. Keep up the great work!
@TheEggMan@jeff007 On-demand is still possible with the help of a hub. That’s how the Blink cameras work, for example. It sounds like the outdoor Wyze cameras will also have a hub of some sort. The hub is always connected to a power source indoors, and it’s always connected to wifi. The cameras are NOT always connected to wifi – they only connect when an event is recorded or when live view is requested. The hub just sends “wake-up” signals to the cameras, basically.
