POE inductively coupled through glass?

stevenkan · Apr 16, 2024

This kickstarter is promising to "deliver 10 Watts through up to 30mm of glass, wood, or anything non-metallic"

In theory it should be possible to couple a POE link through glass as well.

I'd like to run some POE IP cameras outside of a rented building where I don't want to drill through walls.

Has anyone heard of such a thing yet? Google knows nothing.

Paladin · Apr 16, 2024

I've seen Qi wireless power delivery though thin wood and other stuff so glass would be fine. Basically for a window you just need big enough coils and enough power to work over the distance between them (due to double pane windows, etc.). I wouldn't go for a kickstarter for something that is basically already available unless you like the specific implementation and the price seems decent.

Lord Evermore · Apr 16, 2024

You have to make sure it's completely plain glass. Any kind of coating (like most new windows) will probably make it either not work or greatly reduce the power. This is probably a very wasteful transmission technology. Would be interesting to see a good watt-meter put on it and see how much power the indoor module uses to get 10W to the other side. I assume it's a constant output from the transmitter as well, regardless of the actual draw on the outer module, so it's always sucking the maximum power. Or perhaps they're "intelligent" and the receiver tells the transmitter to decrease output/draw if it's not needed to get through a particular material and still output 10W.

I don't know how you'd get PoE power out of this device since it's only got USB output; a quick search only finds devices that do the reverse, and standard PoE requires negotiation between the source and the device so it would be quite an expensive device just to supply a single device with less than 10W. Passive PoE could work but any PoE is going to need voltage conversion. After losses from the transmission and all the conversion, you might not even have enough to power a camera. (Of course, there's no reason to make it PoE cameras instead of USB-powered, since there won't be any data transfer over the connection.)

It would probably be easier and more reliable and efficient to just find routes where you could run Cat5 cabling through existing openings and tack it up along the walls/eaves/siding, and power it from a switch. Flat cables are available that could probably be slipped under the window and still allow it to close and latch, or perhaps they could be run up to the attic and out through the vents around the eaves or a fan opening. Or just use rechargeable cameras.

Paladin said:
I've seen Qi wireless power delivery though thin wood and other stuff so glass would be fine. Basically for a window you just need big enough coils and enough power to work over the distance between them (due to double pane windows, etc.). I wouldn't go for a kickstarter for something that is basically already available unless you like the specific implementation and the price seems decent.

A quick search doesn't seem to find any existing out of the box solutions like this. It is kind of niche, and $65 bucks for what amounts to a very short power cord and low power wall wart is a lot (and the retail price would be a lot more). There are parts for making your own wireless induction, but that also all seems to be meant for much thinner materials, nearly direct contact. This device must be really cranking out the power (relatively) to get through such a distance and probably has to be perfectly aligned.

Andrewcw · Apr 16, 2024

The device is completely viable. But i would be curious what the long term heat will do even though it is very low heat. Being outdoor what will grow or be attracted to this heat source.

But being on kickstarter. I'll wait till it gets released to public. As the outcomes are more likely it is
1) Vaporware
or
2) They found it on Ali/baba/express.

The only thing this is doing is making the receiving end waterproof. You can buy Qi Power receivers off the shelf.

Lord Evermore · Apr 16, 2024

Andrewcw said:
But i would be curious what the long term heat will do even though it is very low heat.

I had my car's windshield glass replaced a couple of years ago by Safelite. My car has a blinking light at the front edge of the dash to indicate the alarm is armed. Just a little red LED that blinks on for 1 second out of 3. I noticed recently that there is a discolored spot on the glass above the light from this dinky little LED that is only on for 1/3 of the time. (The original glass as well as a prior replacement didn't develop this.) This wireless power device could cause similar problems with coated glass windows, or other materials. I don't know if it might do anything to truly plain glass.

Even if they aren't making anything truly innovative, just repackaging something you can already buy (probably not Qi because then someone could find that out eventually), at the very least they're making something it doesn't seem like anybody else is making, and they are doing the work of putting it all together and soldering the connectors on and making the housing and assembling it.

Could be something like this, except this set is MUCH lower power. https://www.adafruit.com/product/1407 Maybe this Power Mole uses all the same basics but they found actually are manufacturing the boards with additional/better components to allow for higher power, or they just found the exact kit they needed and stuck it in the housings.

Kyuu · Apr 16, 2024

Yeah that Kickstarter product is strictly for devices that utilize 5V USB power. It's not doing PoE. This sort of device would likely not be viable for PoE for one simple reason: the whole point of PoE is to send power on the same wire as the data. I suppose it's possible to maybe inductively couple the data signaling as well, but almost certainly not practical.

For your rented building, if you really want to go with something like this, there are network cameras that use 5V USB to power the camera, and then use wifi for the data. That's what the Eufy camera I have over my front door does, although I supplied USB power to the camera from an outdoor light fixture via this Wyze adapter.

Lord Evermore · Apr 16, 2024

Kyuu said:
I suppose it's possible to maybe inductively couple the data signaling as well, but almost certainly not practical.

Just need a Wi-Fi-capable camera, regardless of power. Basic cameras are pretty low power so USB is sufficient (even a motorized camera would easily work with USB-C PD now), but it has a 15-foot limit, although with power-only maybe that's longer. There could be some instances where you can't run data or video cabling all the way to the camera or the location of the power outlet and need it to be Wi-Fi (or proprietary wireless), but it still needs wired power, so a PoE line over 15 feet without data would be used. I can't really think of much reason for it to not just have a power plug and adapter in most cases, other than a PoE port making it compatible with both wired and wireless connections. But these would all be very limited instances.

Placing a security camera where the power source for it is directly exposed and could easily be disconnected doesn't seem like super-great security. Yes, you'd get an initial video, perhaps an alert from monitoring, but that may not be much use. Thirty seconds of covering the face to pull the plug and the thief is safe.

redleader · Apr 16, 2024

stevenkan said:
This kickstarter is promising to "deliver 10 Watts through up to 30mm of glass, wood, or anything non-metallic"

View attachment 78673

In theory it should be possible to couple a POE link through glass as well.

I thought about making one of these for a POE camera a couple years ago, but in the end hired an electrician to snake an ethernet cable. No reason at all why it wouldn't work though, its just a wireless phone charger or wireless toothbrush charger.

Lord Evermore said:
You have to make sure it's completely plain glass. Any kind of coating (like most new windows) will probably make it either not work or greatly reduce the power.

I think it would be ok. Obviously any dielectric low e coating wouldn't interact with a magnetic field. Silver would, so a lot would interact, but to be transparent to light the silver coating has to be nanometers thick and therefore not very conductive. Too lazy to look up the equations, but I suspect it wouldn't actually be a problem in practice just because the material thickness is too low to absorb much.

Lord Evermore · Apr 16, 2024

redleader said:
I thought about making one of these for a POE camera a couple years ago, but in the end hired an electrician to snake an ethernet cable. No reason at all why it wouldn't work though, its just a wireless phone charger or wireless toothbrush charger.

Seems like there would be an insane amount of loss trying to get the power level up to PoE standards, even at 24V, and as I mentioned, if it's not passive PoE, you would need to integrate all the intelligence required for that into the receiver module. So very niche that it's not likely to be worth making it as a product to sell. And I don't know enough about wireless power; does increasing the voltage mean increasing the magnetic field modulation frequency or just the power of it and letting the receiver handle the conversion? And would changing that make it harder or easier to get through glass or other materials? Most wireless chargers are already not the highest efficiency, and they're intended to only go a couple of millimeters through nothing but air but they can provide hundreds of watts now, so it seems like the Power Mole could even be one of those but in order to get through 30mm of glass it's down to a few percent of efficiency just to provide 10W.

stevenkan · Apr 17, 2024

redleader said:
I thought about making one of these for a POE camera a couple years ago, but in the end hired an electrician to snake an ethernet cable. No reason at all why it wouldn't work though, its just a wireless phone charger or wireless toothbrush charger.

This could be your $billion idea!

As I'm thinking about the data part, I wonder if optical would be better than inductive, since it's going through glass anyway.

As far applications go, I'm thinking about something like this camera that's popular for license plate recognition (LPR):

With motorized optical zoom, this is not something that's easily replaced by a WiFi camera, nor anything powered by USB. It draws somewhere between 4.6 - 12.1 W over POE.

Lord Evermore said:
It is kind of niche, and $65 bucks for what amounts to a very short power cord and low power wall wart is a lot (and the retail price would be a lot more).

I would pay $200 for a device like this, if one existed. It would cost more a lot more than that to get a contractor out here to run a cable.

w00key · Apr 17, 2024

redleader said:
I think it would be ok. Obviously any dielectric low e coating wouldn't interact with a magnetic field. Silver would, so a lot would interact, but to be transparent to light the silver coating has to be nanometers thick and therefore not very conductive.

Eeh, 20yo coated argon filled double glass here and it reduces 2.4/5 Ghz power level by 20 dB. It's easier for radio to go through a brick wall than the windows when I had a point to point link running between two houses right opposite to each other using 2x Ubiquiti Nanobeam AC. Going through the window, signal was as if we were km+ apart.

This can't be good if you want to transfer power through it.

KD5MDK · Apr 17, 2024

Lord Evermore said:
(Of course, there's no reason to make it PoE cameras instead of USB-powered, since there won't be any data transfer over the connection.)

Do you have any 100ft USB cables?

w00key · Apr 17, 2024

KD5MDK said:
Do you have any 100ft USB cables?

Should even work up to 300ft. That's the power of 57V (vs 5)

steelghost · Apr 17, 2024

w00key said:
Eeh, 20yo coated argon filled double glass here and it reduces 2.4/5 Ghz power level by 20 dB. It's easier for radio to go through a brick wall than the windows when I had a point to point link running between two houses right opposite to each other using 2x Ubiquiti Nanobeam AC. Going through the window, signal was as if we were km+ apart.

This can't be good if you want to transfer power through it.

Radio waves aren't magnetic fields though....

w00key · Apr 17, 2024

steelghost said:
Radio waves aren't magnetic fields though....

Not sure what wifi you use but mine is in the class of electromagnetic waves. Sure, higher frequency than 87 to 205 kHz used for Qi but still the same physics - it's not physical movement like (ultra)sound or photons at much higher frequencies. Any metal layer attenuates it.

Unlike the 9999 papers describing how it kills waves at 900 Mhz - tens of Ghz, there doesn't seem to be much research at these low frequencies except one paper that tested coated glass for EMP defense:

The conductive-coated glass samples had 4.0064×103∼4.7438×103 (S/cm) conductivity (𝜎); the resistivity (𝜌) was reduced down to 2.1080×10−4(Ω·cm); and the light transmittance was 74% or higher. At a frequency band of 20 kHz∼100 MHz, the glass samples had the minimum SE of 40 dB and the maximum SE of over 80 dB. At a frequency band of 10 kHz, where shielding is challenging, the SE remained at 31.1 dB or higher. At a frequency band of 100 MHz∼1 GHz, the shielding attenuation decreased gradually; however, SE of at least 32 dB was achieved. These glass materials have practical value as high-transmittance conducive materials for windows of automobiles, vessels, and aircrafts to protect against EMP.

Study on Shielding Effectiveness of High Transmittance Coating Film Glasses against Electromagnetic Pulse

This study investigated the shielding effectiveness (SE) of glass materials with conductive coatings by establishing a 3000 × 3000 × 3000 mm electromagnetic pulse (EMP)—shielded room according to the EMP shielding requirements in the US military standard MIL-STD-188-125-1. The EMP SE of...

www.mdpi.com

Minimum shielding effectiveness of 31 dB is still more than plenty to kill any charging attempt. Even 10 dB attenuation is x10 effort needed to get power to move through it already.

Lord Evermore · Apr 17, 2024

KD5MDK said:
Do you have any 100ft USB cables?

This device isn't going to make the connection go 100 feet, so I assumed that wasn't a requirement in this case.

Lord Evermore · Apr 17, 2024

stevenkan said:
I would pay $200 for a device like this, if one existed. It would cost more a lot more than that to get a contractor out here to run a cable.

Sure, if you were going to run the cable all the way from the network closet to the camera, through the walls or ceiling. But that's an entirely different class of connection. THIS device isn't going to give you a setup that will run a camera that requires PoE, so that's not a valid comparison. This is going to give you less than 20 feet of connectivity (from the power outlet) for a very low power camera that could also be accomplished with a drill and a USB charging cable or maybe just a flat cable under a window. That makes THIS device not worth $200.

steelghost · Apr 17, 2024

w00key said:
Not sure what wifi you use but mine is in the class of electromagnetic waves. Sure, higher frequency than 87 to 205 kHz used for Qi but still the same physics - it's not physical movement like (ultra)sound or photons at much higher frequencies. Any metal layer attenuates it.

Qi is a form of inductive charging which is based on oscillating magnetic fields. Totally different physics to WiFi.

Now, for all I know the ultra-thin metal layers on an ordinary modern window might well interfere with rapidly oscilating magnetic fields - but on the other hand, possibly not. But again, not the same physics as radio waves etc.

BTW the Qi standard is apparently able to tranmit power over distances of up to 40mm, although it doesn't mention how much power, and through what...:

The Qi standard has been developed by the Wireless Power Consortium and is applicable for electrical power transfer over distances of up to 40 millimetres (1.6 inches).

w00key · Apr 17, 2024

Inductive charging is still https://en.wikipedia.org/wiki/Electromagnetic_radiation. Exactly same physics as wifi, send current through a wire and a magnetic field form around it. Use a special shape to encourage power to escape, like a dipole antenna, or coils for induction charging or cooking.

All of these get attenuated by non vacuum and reflection, absorption messes it up. The linked paper tested from 10 khz up and coated windows kills all of it.

[edit] Think of it like on a induction cooker - a nice big coil that induces current in a target and heat it up. What happens if you put a sheet of aluminium foil on it? It melts and makes a mess as it catches enough energy even as a poor induction material and if you're unlucky, messes up the cooktop glass.

It doesn't matter that with induction charging you have a source and target coil and nice coupling, stuff a metallic coated thing in between them and the target is swapped out for the coating. The target might as well not be there.

Induction "magnetic" waves are nothing but super long waves, there is no lower frequency limit where electromagnetic waves stop acting like one. Or high limit - gamma photons are also wave-y. While attenuation differs hugely between frequency, paper cites a 30 - 80 dB range which is 10^5 difference in shielding, a metallic coating attenuates pretty much everything and even a few dB is deadly if you don't want to send in 10W for 1W output. It's horrible enough already with Qi, nothing in between and perfect alignment.

redleader · Apr 17, 2024

w00key said:
Eeh, 20yo coated argon filled double glass here and it reduces 2.4/5 Ghz power level by 20 dB. It's easier for radio to go through a brick wall than the windows when I had a point to point link running between two houses right opposite to each other using 2x Ubiquiti Nanobeam AC. Going through the window, signal was as if we were km+ apart.

This can't be good if you want to transfer power through it.

w00key said:
Not sure what wifi you use but mine is in the class of electromagnetic waves.

Those are actually different things. With wifi you're sending microwaves (an electromagnetic field) at the surface where if it is metalized they're going to reflect off it, similar to how light bounces off of metals, essentially the window makes a pretty good mirror at those frequencies. With two coils that have a thin object sandwiched between them you're sending a purely magnetic field through the object, which will not be reflected like an EM field would be. Rather, if the object is dielectric it'll go through with essentially no loss. If the object is conductive, you'll set up eddy currents in the material which will lead to resistive losses that increase with frequency and the thickness of the object. Since we're talking about low frequencies (probably low tens of kilohertz from the size of the coil) and a layer that is so incredibly thin light can pass right through it with negligible attenuation, you won't induce much current and so losses will be very low.

Lord Evermore said:
Seems like there would be an insane amount of loss trying to get the power level up to PoE standards, even at 24V, and as I mentioned, if it's not passive PoE, you would need to integrate all the intelligence required for that into the receiver module.

Typical wifi camera is like 4 or 5w. Even if you waste 20% of that in the coil windings, you're only out 1 watt or less which is not a big deal. Voltage in this case is trivial, the device is literally a transformer. It can produce any voltage you want. If you want to do active PoE, put a controller. Most likely you don't need it though since cameras will generally take external power as well as PoE.

Lord Evermore · Apr 17, 2024

redleader said:
Typical wifi camera is like 4 or 5w. Even if you waste 20% of that in the coil windings, you're only out 1 watt or less which is not a big deal. Voltage in this case is trivial, the device is literally a transformer. It can produce any voltage you want. If you want to do active PoE, put a controller. Most likely you don't need it though since cameras will generally take external power as well as PoE.

I wasn't sure if cranking up the power to allow for higher voltage would result in higher loss rate trying to send it inductively, especially through any material that might block it. For doing active PoE, it's just that it would result in higher cost for the electronics and design (none of the wireless power kits shown is anything but dumb transmission, so it would also require someone with the knowledge of the electronics needed, not just somebody that can shove off-the-shelf stuff into a plastic casing). But yes, if you specifically were setting up a camera that accepts straight DC power then you could also specifically build or buy a wireless power device like this that puts out DC power, or adapt the USB output of off the shelf models (probably needing to convert 5V to 12V), but in a situation where PoE is what you need, it doesn't seem like it would be cheap to do. A universal transmitter model would probably work, and just swap out the receiver for what you need. They could even make a universal receiver with a swappable module to plug in that converts to whatever output you need but that might be more expensive than just making entire receivers.

w00key · Apr 17, 2024

redleader said:
With two coils that have a thin object sandwiched between them you're sending a purely magnetic field through the object, which will not be reflected like an EM field would be. Rather, if the object is dielectric it'll go through with essentially no loss. If the object is conductive, you'll set up eddy currents in the material which will lead to resistive losses that increase with frequency and the thickness of the object.

[Citation required]

My paper went down to 10 khz and windows block EMP well, a broad spectrum "fuck everything" signal. Pure glass is a dielectric but silver coated glass isn't.

Now your turn, cite something that shows coupled coils ignore a conducting plate placed between them.

redleader · Apr 17, 2024

w00key said:
[Citation required]

My paper went down to 10 khz

A 10 KHz electromagnetic wave is not the same thing as a 10 KHz oscillating magnetic field, so this isn't relevant. An electromagnetic wave is a a propagating stream of RF energy that can travel unlimited distance because the electric and magnetic fields continually reproduce one another. A magnetic field alone doesn't do that, the field lines fold back to the opposite pole of the magnet.

w00key said:
Now your turn, cite something that shows coupled coils ignore a conducting plate placed between them.

"If the object is conductive, you'll set up eddy currents in the material which will lead to resistive losses that increase with frequency and the thickness of the object."

Note that I'm not saying they ignore the conductor, just that the effect depends on frequency and thickness and gets really small for thin materials and low frequencies. You can look this up, there is a whole wikipedia page on eddy currents.

stevenkan · Apr 17, 2024

Lord Evermore said:
This is probably a very wasteful transmission technology. Would be interesting to see a good watt-meter put on it and see how much power the indoor module uses to get 10W to the other side. I assume it's a constant output from the transmitter as well, regardless of the actual draw on the outer module, so it's always sucking the maximum power. Or perhaps they're "intelligent" and the receiver tells the transmitter to decrease output/draw if it's not needed to get through a particular material and still output 10W.

From the Power Mole website:

we could have made a cheap product where the transmitter simply cranks out magnetic fields continuously. Besides being wasteful of energy, these fields can heat up metal objects in the vicinity.

To account for this, we made the transmitter and receiver capable of communicating with each other. If the transmitter cannot find a receiver, the magnetic fields turn off. If the receiver is drawing too much power, the transmitter will also stop the power transfer. And in case a metal object is between the transmitter and receiver, the transmitter also monitors the input and output power levels to ensure stray power is not being significantly absorbed by nearby objects. This feature ensures nearby metallic objects are not heated up by the transmitter.

Lord Evermore said:
I don't know how you'd get PoE power out of this device since it's only got USB output; a quick search only finds devices that do the reverse, and standard PoE requires negotiation between the source and the device so it would be quite an expensive device just to supply a single device with less than 10W. Passive PoE could work but any PoE is going to need voltage conversion. After losses from the transmission and all the conversion, you might not even have enough to power a camera. (Of course, there's no reason to make it PoE cameras instead of USB-powered, since there won't be any data transfer over the connection.)

Again, I'm not suggesting that I use the Power Mole. I'm suggesting that, if the Power Mole is technically feasible, it might also be feasible to build a "Power Mole POE Edition" that does true POE using similar power transmission plus some data-transmission-technique TBD-by-someone-smarter-than-me.

Lord Evermore said:
It would probably be easier and more reliable and efficient to just find routes where you could run Cat5 cabling through existing openings and tack it up along the walls/eaves/siding, and power it from a switch. Flat cables are available that could probably be slipped under the window and still allow it to close and latch, or perhaps they could be run up to the attic and out through the vents around the eaves or a fan opening. Or just use rechargeable cameras.

A quick search doesn't seem to find any existing out of the box solutions like this. It is kind of niche, and $65 bucks for what amounts to a very short power cord and low power wall wart is a lot (and the retail price would be a lot more). There are parts for making your own wireless induction, but that also all seems to be meant for much thinner materials, nearly direct contact. This device must be really cranking out the power (relatively) to get through such a distance and probably has to be perfectly aligned.

The type of camera I want to use is linked above, and does not lend itself to USB or battery power.

And this is the room I'm thinking of at this very moment. None of the windows opens. I had Cat5e run to a wall jack on the wall to the left of the exterior door. I want to mount a couple of those cameras for license plate monitoring, outside on that balcony. Right now I have a ghetto cable running over the top of the door, but obviously that's not a viable long-term solution. The underside of that red awning thing is metal, so the landlord is loathe to let me drill through it. The terms of my lease are also such that I need to undo any changes I make to the building when my lease ends, and I can't use drywall patch on metal.

Kyuu · Apr 17, 2024

Monitoring license plates? Seems sketchy.

But in any case, I wouldn't hold your breath for a solution that will let you operate a PoE camera without running a cable to it. It is possible to patch small holes in metal (basically the same as repairing a small ding in a car with Bondo). You can also drill through some other part of the exterior and then just do a surface run using adhesive fasteners/channel.

Paladin · Apr 17, 2024

We had/have it at my workplace. Basically its a small business park and for reasons I still don't understand, the park ownership/management wanted to track license plates coming and going... as if they could do anything with the information? They bought all the cameras, spent a bucket of money getting them installed, calibrated and setting up the frankly terrible software to do the camera recording and recognition stuff. A server just for that actually, at first.

And it never worked all that well. Without cameras right at the perfect spot they just never worked all that well and given that the parking is all outdoors, the lighting was almost always bad, the distance and angles were wrong most of the time and the few times it did produce an accurate result... so what? It's an open lot so they don't know who the plates belong to and never got accurate enough to be able to restrict access or require anyone to register their plates, etc. They couldn't bar any delivery vehicles or guests or anything anyway so there was no point.

Even with all that, it was really disappointing. Even under near ideal conditions, the system never worked as well as the sales people promised. The cameras generated a TON of data/video on the recording server, they are 2k cameras with low light/night mode etc. but it was never even close to good enough.

I imagine if you had indoor parking with a perfect setup where the plates are always moving slowly toward or away from a camera rig right at the entrance/exit, and lighting is always good... maybe it would work. Even then, you will still need more to actually control access effectively so it's kind of pointless.

Lord Evermore · Apr 17, 2024

To account for this, we made the transmitter and receiver capable of communicating with each other. If the transmitter cannot find a receiver, the magnetic fields turn off. If the receiver is drawing too much power, the transmitter will also stop the power transfer. And in case a metal object is between the transmitter and receiver, the transmitter also monitors the input and output power levels to ensure stray power is not being significantly absorbed by nearby objects.

I imagine this was just part of the off-the-shelf parts, and not really involving a lot of communication. It's just an on-off trigger based on thresholds; if the receiver doesn't get a strong enough magnetic field, it flips something that the transmitter detects and the transmitter turns off. Maybe as simple as something like the security stickers on retail products. If that doesn't receive enough power, the transmitter can't read it and assumes there is a problem. The receiver can also have a way to drain the power from it so it stops responding. There's no indication that it really controls the levels of the transmission intelligently, varying the output from the transmitter based on the needs of the device that is plugged into the receiver, or really "monitoring" the power levels on the transmitter side.

Plenty of gap around that door for a flat Cat5 cable it looks like. Or maybe you could lift up the edge of the outside awning metal to pass a cable through and inside above the ceiling tiles, if there's a small opening, using a fishing tool. Or are there any light fixtures out there, so perhaps there is conduit (which again might pass through a hole where you could pass a cable, if going inside the conduit isn't feasible). Or just punch through the area above the frame of the glass and out, if it's not metal on the outside face. I can't tell how far down the part with the slats goes. At the very least, take a look above the tiles. Maybe you'll see something that makes it all surprisingly easy. Rental properties like that are always shoddily-built.

Lord Evermore · Apr 17, 2024

Paladin said:
We had/have it at my workplace. Basically its a small business park and for reasons I still don't understand, the park ownership/management wanted to track license plates coming and going... as if they could do anything with the information?

Probably some consultant convinced them that they could do some sort of marketing based on the information. Somehow. Seeing how often the same people come and go, and when, how long they stay. Maybe it would be correlated with what companies are leasing the space in the building and they'd be able to see which types attracted the most unique visitors. And if it was accurate and clear, they could identify vehicles that sit there 9 hours every single day and therefore know how many spaces were being taken up by employees and whether they should start charging extra rent or requiring registration or something. Or identify people using the lot constantly but not patronizing the businesses or working there. It might not have actually been useful to the ownership, but they were clearly convinced it would be by the salespeople. Anything that they're told will save money or let them charge more can be sold to management. Or maybe just plain security, being able to identify vehicles that are there with a unique identifier that has records associated, rather than just comparing appearances.

Lord Evermore · Apr 17, 2024

You know, depending on the cost and just how important this is, just have someone come in and remove the small pane of glass over the door and put in one with a matching frame that leaves a section where you could put in other material that would let you run the cable, without being so large that it's a security problem. Save the original and put it all back when you leave. Or if you had cable run to the left of the door, is the outside of that wall something you couldn't repair when you leave? (I don't see the "ghetto cable" but unless it might get damaged or you're really particular about the appearance, if it works, leave it.)

I just think there has to be some opening where a cable could be passed through either without damage or in material that you could repair later.

Paladin · Apr 17, 2024

Lord Evermore said:
Probably some consultant convinced them that they could do some sort of marketing based on the information. Somehow. Seeing how often the same people come and go, and when, how long they stay. Maybe it would be correlated with what companies are leasing the space in the building and they'd be able to see which types attracted the most unique visitors. And if it was accurate and clear, they could identify vehicles that sit there 9 hours every single day and therefore know how many spaces were being taken up by employees and whether they should start charging extra rent or requiring registration or something. Or identify people using the lot constantly but not patronizing the businesses or working there. It might not have actually been useful to the ownership, but they were clearly convinced it would be by the salespeople. Anything that they're told will save money or let them charge more can be sold to management. Or maybe just plain security, being able to identify vehicles that are there with a unique identifier that has records associated, rather than just comparing appearances.

Possibly yeah but I wish I had been in that meeting. It's a small office business park where tiny companies like 2 person insurance and law offices lease space, small team customer service centers etc. The only real concern is basic security. No sales type stuff that could benefit from marketing other than the office spaces themselves. I really don't know what they thought they would get out of it. There was never any kind of talk of locking down access to the parking area or anything. It would have been a crazy amount of work to try to wall it all off or something. You can just walk into the lot anywhere.

I think it must have been something about trying to just relate which cars belong to who but they never even started down that road, even when they first thought it was going to work. You would think they would have either moved ahead if it seemed viable or quickly given up when the consultant could not make it reliably scan any plates. Instead it was left to linger for years and now I think they just use them as regular old IP cameras, which we already had like 30 of.

redleader · Apr 17, 2024

Kyuu said:
But in any case, I wouldn't hold your breath for a solution that will let you operate a PoE camera without running a cable to it. It is possible to patch small holes in metal (basically the same as repairing a small ding in a car with Bondo). You can also drill through some other part of the exterior and then just do a surface run using adhesive fasteners/channel.

Yeah that was my take as well. I thought about it for a bit then just had someone run a cable. Still, I spent a few minutes on Amazon just now and there are some surprisingly interesting wireless power devices:

View: https://www.amazon.com/Taidacent-Distance-Inductive-Charging-Wireless/dp/B07VZZR4ZK/ref=sr_1_5

This (bad!) review mentions:

I was able to get 12V at about 1.5A out of it at 2" distance.

That is a surprising amount of power through that distance

We have a fair amount of wildlife (coyotes, foxes, deer) wandering through our yard, it would be neat to be able to clamp a 4k camera on any window for a couple days at a time, but I should probably just buy a battery powered trail cam and stick on a tree.

Paladin said:
We had/have it at my workplace. Basically its a small business park and for reasons I still don't understand, the park ownership/management wanted to track license plates coming and going... as if they could do anything with the information? They bought all the cameras, spent a bucket of money getting them installed, calibrated and setting up the frankly terrible software to do the camera recording and recognition stuff. A server just for that actually, at first.

And it never worked all that well. Without cameras right at the perfect spot they just never worked all that well and given that the parking is all outdoors, the lighting was almost always bad, the distance and angles were wrong most of the time and the few times it did produce an accurate result... so what? It's an open lot so they don't know who the plates belong to and never got accurate enough to be able to restrict access or require anyone to register their plates, etc. They couldn't bar any delivery vehicles or guests or anything anyway so there was no point.

Even with all that, it was really disappointing. Even under near ideal conditions, the system never worked as well as the sales people promised. The cameras generated a TON of data/video on the recording server, they are 2k cameras with low light/night mode etc. but it was never even close to good enough.

I imagine if you had indoor parking with a perfect setup where the plates are always moving slowly toward or away from a camera rig right at the entrance/exit, and lighting is always good... maybe it would work. Even then, you will still need more to actually control access effectively so it's kind of pointless.

If you look on google there are whole forums of people who do this kind of stuff, and there are community favorite models for different applications (cars, night surveillance, weather photography, etc). If you want to get license plates it is all about having a very carefully lined up camera and a pretty long focal length lens. I wasn't so interested in that, but I ordered one of these random "empiretech" cameras those forums recommended for low light and it is pretty remarkable. It makes the night mode on my Pixel 8 pro look embarrassingly bad and does so without insanely long exposure times. I think if you really pick the right equipment for the right job you can do pretty well, but theres a lot of expertise involved. I'm not surprised some random sales guys did a shit job for your work.

Lord Evermore · Apr 17, 2024

Paladin said:
Instead it was left to linger for years

Really good salesperson/consultant got a solid multi-year contract signed that said nothing about whether it would work or not?

Paladin · Apr 18, 2024

Lord Evermore said:
Really good salesperson/consultant got a solid multi-year contract signed that said nothing about whether it would work or not?

More like they sold it to the park owners with no tech people being involved and then after the implementation failed to live up to their promises, the park owners felt kind of foolish and never really pushed it. Fortunately there was no real ongoing opex beyond the IT time to keep it running and poke it with a stick to try and get more performance from it for a while until everyone got tired of trying to make it lay a golden egg or something. I would say it was a lesson learned but the park ownership people have really short memories for some of this stuff. One of the main issues was that no one was ever really all that clear on what it was supposed to accomplish in the first place, the motive for actually buying it, so no one really felt that bad that it didn't work out.

whoisit · Apr 18, 2024

Lord Evermore said:
I had my car's windshield glass replaced a couple of years ago by Safelite. My car has a blinking light at the front edge of the dash to indicate the alarm is armed. Just a little red LED that blinks on for 1 second out of 3. I noticed recently that there is a discolored spot on the glass above the light from this dinky little LED that is only on for 1/3 of the time. (The original glass as well as a prior replacement didn't develop this.) This wireless power device could cause similar problems with coated glass windows, or other materials. I don't know if it might do anything to truly plain glass.

Interesting. There have been a couple reports of the backplates on MSI video cards showing burn in from the LEDs on gaming ram.

https://www.techspot.com/news/102548-ram-rgb-lights-might-bleach-gpu-backplate.html

Gamers Nexus also has a snippet on it.

Kyuu · Apr 18, 2024

redleader said:
If you look on google there are whole forums of people who do this kind of stuff, and there are community favorite models for different applications (cars, night surveillance, weather photography, etc). If you want to get license plates it is all about having a very carefully lined up camera and a pretty long focal length lens. I wasn't so interested in that, but I ordered one of these random "empiretech" cameras those forums recommended for low light and it is pretty remarkable. It makes the night mode on my Pixel 8 pro look embarrassingly bad and does so without insanely long exposure times. I think if you really pick the right equipment for the right job you can do pretty well, but theres a lot of expertise involved. I'm not surprised some random sales guys did a shit job for your work.

We're veering off-topic a bit here, but even if you can reliably get the license plate data, the question remains: and then what? What do you do with that?

Paladin · Apr 18, 2024

Kyuu said:
We're veering off-topic a bit here, but even if you can reliably get the license plate data, the question remains: and then what? What do you do with that?

Yeah... I can see if you have a facility where you can limit 100% who can drive in to the lot where you have the plate cameras, then I guess it makes sense but... don't you just have them use an access card/code for the gate anyway? I mean, when an employee needs to rent a car or buy a new car etc., do they get locked out until they get things updated through whatever procedure? I'm sure there are some facilities like that where they want it as some kind of factor in security or whatever but it still seems kind of odd. I get how valuable it would be on a police car but for a parking garage on a private property, I don't really see much practical point to it.

stevenkan · Apr 18, 2024

Kyuu said:
We're veering off-topic a bit here, but even if you can reliably get the license plate data, the question remains: and then what? What do you do with that?

My office neighbor had his truck stolen the other day, and after reviewing the footage from my existing cameras, we noticed a suspicious white minivan that appears to have been the thief's "lookout" vehicle. It arrived and parked in the lot about 45 minutes before the thief arrived on a bicyle, did a lap around the back and front parking lots about 10 minutes before the theft, and then left the parking lot about 5 minutes after the theft.

And then they did it again!!!! Van arrives, waits, does a lookout lap. Truck returns to steal another set of tools, and bolts. Van leaves very soon afterward.

None of my existing cameras can capture plates. But if we had plates that might give the police something to go on.

KD5MDK · Apr 18, 2024

Most of the people on IPCamTalk seem convinced that there are hordes of ne'er-do-wells just waiting to break a window and steal their antique silverware or and the police will do something about it if you give them just the right screenshot walking up to the front door.

stevenkan · Apr 18, 2024

Paladin said:
We had/have it at my workplace. Basically its a small business park and for reasons I still don't understand, the park ownership/management wanted to track license plates coming and going

I wish I had your landlord

. I'm on my own, here.

redleader said:
If you look on google there are whole forums of people who do this kind of stuff, and there are community favorite models for different applications (cars, night surveillance, weather photography, etc). If you want to get license plates it is all about having a very carefully lined up camera and a pretty long focal length lens. I wasn't so interested in that, but I ordered one of these random "empiretech" cameras those forums recommended for low light and it is pretty remarkable. It makes the night mode on my Pixel 8 pro look embarrassingly bad and does so without insanely long exposure times. I think if you really pick the right equipment for the right job you can do pretty well, but theres a lot of expertise involved. I'm not surprised some random sales guys did a shit job for your work.

Yes! I just bought an IPC-B52IR-Z12E S2 based on a recommendation from the knowledgeable folks over at IPCamTalk, and it's a pretty amazing camera. As it turns out my location is not great for license plate capture, so I'm going to have to see if my neighbors are willing to host (or buy) a set of these cameras. There are 3 entrances to the complex, so I think 3 cameras would do it.

stevenkan · Apr 18, 2024

KD5MDK said:
Most of the people on IPCamTalk seem convinced that there are hordes of ne'er-do-wells just waiting to break a window and steal their antique silverware or and the police will do something about it if you give them just the right screenshot walking up to the front door.

We had a really scary break-in a few months ago where the guy threatened to kill my employee. Based on video from my inside security camera, which he looked directly at, the police nabbed him a few months later. Prosecution is still ongoing.

POE inductively coupled through glass?

Ars Legatus Legionis

Ars Legatus Legionis

Ars Scholae Palatinae

Ars Legatus Legionis

Ars Scholae Palatinae

Ars Tribunus Militum

Ars Scholae Palatinae

Ars Legatus Legionis

Ars Scholae Palatinae

Ars Legatus Legionis

Ars Praefectus

Ars Legatus Legionis

Ars Praefectus

Ars Praefectus

Ars Praefectus

Ars Scholae Palatinae

Ars Scholae Palatinae

Ars Praefectus

Ars Praefectus

Ars Legatus Legionis

Ars Scholae Palatinae

Ars Praefectus

Ars Legatus Legionis

Ars Legatus Legionis

Ars Tribunus Militum

Ars Legatus Legionis

Ars Scholae Palatinae

Ars Scholae Palatinae

Ars Scholae Palatinae

Ars Legatus Legionis

Ars Legatus Legionis

Ars Scholae Palatinae

Ars Legatus Legionis

Ars Tribunus Angusticlavius

Ars Tribunus Militum

Ars Legatus Legionis

Ars Legatus Legionis

Ars Legatus Legionis

Ars Legatus Legionis

Ars Legatus Legionis