Best strategy for remote access?

[cosmo]

Ha! That's funny. Well, they are gardeners and cleaners, so if they're hanging around surfing the web, they're not working!

But in all seriousness, I get your point and it is a valid one. A mobile connection is pretty poor at this particular property down on the coast and nestled in a low point. I had DSL put in for that very reason. That's a pretty clever strategy. Give everyone the Wi-Fi password and see when a connection is established and lost.

 

[cosmo]

use the PoE switch port with Watchdog features to power the modem

So is this how it would be connected:

- Modem connected to PoE switch.
- Switch has PoE cable going into a a PoE splitter (RJ45 female with PoE splits into a RJ45 male and a 5.5mm x 2.5mm DC male connector)
- DC male connector from switch in turn powers modem. RJ45 male goes nowhere.
- Switch powers off modem for a short period if ping fails.

I've never used a managed switch so I'm just trying to figure out how it is all hooked together to be able to reset a device.

 

[ruppmeister]

So is this how it would be connected:

- Modem connected to PoE switch.
- Switch has PoE cable going into a a PoE splitter (RJ45 female with PoE splits into a RJ45 male and a 5.5mm x 2.5mm DC male connector)
- DC male connector from switch in turn powers modem. RJ45 male goes nowhere.
- Switch powers off modem for a short period if ping fails.

I've never used a managed switch so I'm just trying to figure out how it is all hooked together to be able to reset a device.

Yup, you got it.

 

[nayr]

basically convert your modem to PoE, there are various adapters, etc.. it will depend on your modem, the voltages it runs at and what you got to work with.. DC voltage is super easy to work with..

the ethernet cable can still uplink your switch and modem, while providing power to the modem.


and @cosmo, you could try a z-wave front door lock.. you will be able to open/close the lock via intenet and have a log of everyone who uses it.. and perhaps a few z-wave light switches so you can ensure nobody leaves the lights on, or you have 'activity' timers that randomly simulate someone is actually living there.

 

[cosmo]

That Z-wave stuff is pretty cool. There's a whole bunch of new product lines out there using Wi-Fi, GSM etc. I've been developing them for years.

Gonna dig into the HikVision SDK today and see if I can't develop my own remote monitoring for low bandwidth situations. The Hik stuff is a hodge podge. It really isn't made for remoting and certainly not for low bandwidth on the upload side. It's probably not an issue for cable users but DSL upload in far away lands can be pretty crappy. I have specific requirements. I usually end up rolling my own apps to get what I want. Takes a lot longer of course, but you learn how things really work along the way.

 

[SquareEyes]

Ha! That's funny. Well, they are gardeners and cleaners, so if they're hanging around surfing the web, they're not working!

Security is 80% psychology and ~20% technology. Contractors are trusted to fulfill their end of any agreement whether you are there or not. I endeavor to give people enough rope to prove themselves. Basic security principals such as signage can be a great first line of defense. Hell, the military and high value targets, still use signage as a first line of defense. This works better if there is some technology (cams) that can be demonstrated to staff.

But in all seriousness, I get your point and it is a valid one. A mobile connection is pretty poor at this particular property down on the coast and nestled in a low point. I had DSL put in for that very reason. That's a pretty clever strategy. Give everyone the Wi-Fi password and see when a connection is established and lost.

I like to use fingerprint time clocks but a strategic camera observing a sign in/out time-sheet may be a reasonable use of CCTV. You just need to know what the individual contractors and their staff look like.

 

[nayr]

you seem tech savvy enough for this idea @cosmo, I just put a usb HID prox reader near my front door.. it basically acts as a usb keyboard and anytime it reads a card it types in the card's ID number and hits enter.

I wrote up a quick dirty lil app hooked to my RPI to read the card, check it against a file of good ID numbers, and then enable/disable the alarm system..

You could do something similar for your contractors, have them badge-in and badge-out to access the property and you'll have the logs/notifications, and can have your security system run in a less verbose mode while they are there working, so you dont get 50 alerts while they are mowing the yard.. i bet most already have HID cards in there possession, you just have to read them in and correlate who has what badge.

If you hook it up to a door lock you can give those badges out instead of keys, then you can remotely fire any of your contractors and just revoke there card without needing to re-key... would be great too if you ever decide to sublet out your property, ie, vacation rental it from time to time.. especially if there is some big event nearby and you can make big bank off it just a few days a year.

 

[cosmo]

After reverse engineering HID readers for a project which is currently being prototyped on city vehicles, I feel like I have the key to the city.

But, as Forrest Gump would say, "That's all I have to say about that."

 

[cosmo]

We also prototyped a simple Arduino/Bluetooth driven door striker 3 years ago, proximity triggered by BLE from an Android mobile and a sprung door. You walk up to your door and it just opens. Like out of the Adams Family. We thought it was great. But it creeped out girlfriends.

 

[nayr]

the hid readers just output a pain text to the access control system, there is nothing secure about that.. the security lies in.. make a card that that hid reader will read with my id on it.. now there's the trick.

reading em is easy, emulating one is hard.. because the reader does the crypto handshake, its all transparent like its supposed to be

crack these cards and you can get in to anything, but they havent been cracked yet.. some of the early ones have been but Ive not seen any of those in eons

most of the time you configure the card reader to output the ID with a prefix and postfix, consider this a seed.. if you unplug my card reader and plug in a USB keyboard and type in the ID from my key.. it will fail, because it will not have the prefix, delimeter, and postfix I have programmed into the reader to wrap around my ID.. and it checks for the whole thing... the config is written one way and not read so if you unplug the reader you cant easily figure out what its config is.

my program is also looking for the specific USB ID and SN of the prox reader and not just any keyboard, so that could slow you down a bit too.. by then the 30's timeout is up and the alarm is going off.

for enabling and disabling my alarm system its adequate, I have a pin-key to get inside and no key.. so for 2 factor I have the thing I know (8-digit pin to unlock) and the thing I have (hid fob to disarm).. ass backwards but I dont like keys, cant revoke em.. its just a residential front door, so I just have to be more secure than those around me

Went with Z-Wave because it provides a lil bit of crypto at least, and it still wont disarm the alarm system so I consider it more difficult than just picking the deadbolt.. really like having the dead bolt auto-lock, know nobody leaves things unsecured.. and with keypad nobody gets locked out in there jammies in winter.

 

[SquareEyes]

Cosmo have you considered an Arduino Uno teamed up with a finger print reader? Should be less than $50 all up on Ali.

Personally I wouldn't use an card/token system as anyone can tag on or off.

 

[cosmo]

Currently we just hide a key in the garden. Works well.

 

[cosmo]

Arduino projects end up being quite a bit more expensive than the board. Here's some photo of one of the units we built with an HID reader that's installed in a City of San Francisco hybrid truck. The board was the cheapest part. Shields can add up. We built an assembly jig to the initial run could be hand soldered and assembled in around 20 minutes. It gets to a point where you just get a board hand made by local companies like Gener8, then it ends up being a lot cheaper. But to prototype, sure, Arduino is fine. But limited. Uno is very limited on memory. So we went to the Mega, which is a bit better. At some point you start looking at RP.

 

[CamFan]

Here I use an outdoor PIR and it's wired to my camera. Milestone emails me 3 photos every time it goes off AND the alarm is set.

PIR gets me a few false alarms a few times a year, hot and cloudy days are tough. Took a while to dial in the settings as it's a fine line between false alarms and not missing cars. Or people.

If it's only vehicles you are looking to detect, look at Cartel driveway sensors, no false alarms.

 

[cosmo]

Interesting. I will look into this alternative. Thanks.

The vision I had a month ago for how the system would work is evolving based upon my experience so far with the real life performance of IR, motion and VCA settings like line crossing. Coupled with how I am finding the software really works remotely and the great feedback and suggestions I am getting from this forum, the way I end up doing this will most likely be quite different to what I envisaged.

It all started with two quotes ($4k and $7k) to supply and install a 4 camera NVR system, which I originally rejected as well padded. After road testing a couple of systems, I have found it is one thing to plug it in and have it record 24x7 with low precision motion detection. Any nong could get that working out of the box. It is entirely different to have a robust, hack resistant and effective surveillance system with high precision notifications that can be viewed remotely. That requires quite a bit more thought and experimentation. And it probably won't be an off the shelf solution from one vendor. More likely a couple of things linked together.

At least hardware in 2016 is pretty cheap compared to 10 years ago.