Alarm System Advice and Next Steps


I’m looking to set up an alarm system built around Z-Wave devices with openHAB acting as the controller. Our current setup is using a SmartThings hub, but I’m trying to move away from that (ew, Samsung).
I currently have some Z-Wave motion sensors, sirens, and door sensors that I’d like to add to the openHAB system, but I’m missing a good keypad solution as well as other core alarm requirements like battery backup, 3G/LTE backup, etc. I’m open to most paths, even if I haven’t listed them here.

  1. Use openHAB as the core alarm system and build features such as SMS, 3G/LTE fallback, battery backup, etc. I’m not sure whether openHAB supports this functionality or how reliable this may be. I’m also not sure how I’d integrate a touchpad.

  2. Use openHAB in conjunction with an EnvisaLink 4, panel, etc. to get a keypad, and bind my Z-Wave sensors to a panel. I’m not sure which (if any) panel will allow random Z-Wave sensors to join the network, but I imagine that if there is a solution like this, then it would likely be the best path forward. I’d also want to build SMS, 3G/LTE fallback into the system.

  3. Use a combination of openHAB and a panel to link Z-Wave sensors to a panel network.

I’m really trying to determine whether it’s even possible to mix Z-Wave sensors with other systems to handle alarm functions. Ultimately, I’d harden the system as much as possible to ensure reliability without having to pay exorbitant fees to ADT and the like.

Thanks for your help!

Well. The problem is and will continue to be that alarm systems aren’t built to allow for use with 3rd party SHARED sensors and subsystems (from their pov) or to integrate with a “master” control system such as openHAB. The question is where to execute the alarm logic. The entity to do that needs full direct access to all sensors (at least those to use for alarm purposes) there.
The mix-and-match variant you’re (understandably) looking for: sorry but forget about it. You won’t get it to work reliably without dedicating sensors to alarming. Feel free to die trying :slight_smile: but I wouldn’t hope for and save those efforts and disappointments right away.

That’s leaving you with one of two choices:

  1. either get a more-or-less standalone alarm system and attach dedicated sensors.
    Usually these are available from the alarm system ecosystem and usually wired, possibly wireless but proprietary in that case, too. There might be systems to allow for ZWave sensors but I doubt there are.
    Either way, point is these sensors will always remain to be exclusive for use with the alarm system and not visible to openHAB for reuse for non-alarm purposes. In effect, you’ll end up having multiple sensors in every place where one should be sufficient, one for alarming and one for smarthoming.
  2. do it all in openHAB including the alarm programming. Yes in technical terms openHAB can do it all for you. Your smarthome server should be running on battery backup anyway, and there’s many messaging interfaces.
    Downside obviously is you need to do it yourself and it’ll be as-reliable as are your programming arts. But you have the full freedom of implementation.
    You wouldn’t even need to integrate a keypad (although you can, of course).
    Btw, this is what I did myself albeit I must admit I still haven’t activated the system, mainly because there have been various reliability problems with wireless sensors and other parts of the system which would have caused false alarms.

And if I wanted to rod you I would just need one of these:

I would NEVER rely on wireless for security related sensors

Well that’s a different question: how you rank security, how likely you are a target to high tech burglars
(they cannot be sure you’re EXCLUSIVELY relying on wireless comms for alarming) etc.
I think the OPs idea is to just make 2nd use of what’s already there, and there’s nothing wrong with that.

No of course there isn’t.
But in my area, the south east of england, around London, there have been an increasing trend for burglaries carried out with these devices. All the burglars need to do is carry one of these and turn it on. They are quite cheap on ebay and more reliable than the sensors themselves.
RF has it’s drawbacks when it comes to reliability mostly and in that case in a security point of view I would rather have my reliability as close to 100% as possible.

With all this in mind, has anyone written a binding for a jammer detector? :wink:

1 Like

I personally haven’t researched this much but I do know that the potential penalties for using such a device in the US would be quite severe. It would make a simple burglary which is handled at the local level into a federal offense. As a criminal, you REALLY don’t want to deal with the federal courts.

I’m not certain that simple possession of such of a device is illegal, but using it would be a violation of federal law and will be very expensive, punishable with steep fines and potential jail time. Using it in the commission of a crime makes the punishment really severe.

So, in the US at least, as a criminal I would either have to be ignorant (which isn’t outside the realm of the possible) or I would not risk taking something that would be a few years in state prison or jail and turn it into 10-15 in a federal prison.

Markus’s also has a good point. The jammer would only be effective if the criminals know you are using wireless sensors and exclusively wireless sensors. At that point, they are a motivated threat and will likely find a way past your sensors either way. On the other hand, if someone puts a SimpliSafe sign in their window, then they do know that your security system is wireless.

I’m not saying that jammers can’t be used. But I don’t think it is as simple as the average corner thug carrying this thing around and now they know they are safe. It would be an interesting research project. I’ve not seen anything in the press and trade publications that I follow that indicate it is a serious threat in the US yet. It is interesting that it is a problem in England.

Now, what IS a threat in the US that I’ve read a vit about are signal boosters for keyless entry keyfobs. They come by your house, boost the signal from your keyfob in the house or in your pocket, unlock the door and drive away.

It will never be 100% reliable so such a goal is somewhat self-defeating. What is a better goal is to provide a system with mitigations for when there is a failure. You will always have failures and you will always have to deal with them. So you need a fallback mitigation anyway. If your mitigation is good enough, it doesn’t matter if your sensors are only 90% reliable.

For example, a mitigation could be if enough sensors go offline simply treat that as an alarm event itself. The biggest challenge with that is avoiding too many false alarms for the usual reasons. Another mitigation can be putting valuables in a safe, insurance that covers theft, etc.

Note that using wired sensors is a mitigation for the potential for wireless sensors to be jammed. But wired sensors are expensive to install so the cost needs to me weighted against the actual risk posed by jammers. For you Vincent, the risk is somewhat high. For me, the risk is pretty low.

Risk = likelihood a threat will exploit a vulnerability * impact if they are successful

Mitigations are ways to further reduce the impact and impact is best represented as money.

The total cost of your mitigations must be less than the Risk itself or else you are wasting money.

In your environment, the Risk is likely high enough that it is worth the cost of using wireless.

In my environment, the suburbs of a city in the US with one of the lowest crime rates where the use of these types of jammers is highly illegal the Risk is pretty low. Everyone needs to assess the Risk based on their particular environment.

That’s an interesting idea. I wonder if one could use an SDR to look for a strong signal in the jammer ranges you care about. If the signal is strong enough you can assume it is a jammer and send an alert (wired obviously :wink: ). @luckymallari, you have some experience with SDR, would something like this be feasible? I know just enough about radios to be dangerous but it seems to me we wouldn’t need to decode any actual information, just look for powerful signals in certain bands or frequencies.

I actually don’t mind dedicating the sensors to the alarm. I don’t really automate very much other than running a few timers, so losing that functionality isn’t a big deal. Any idea where I can find an alarm controller that follows the z-wave spec?

My comment was a bit tongue-in-cheek but you don’t actually need a binding. You can buy an off-the-shelf jammer detector with a NC/NO relay output so it would be trivial to connect this to a RPI or similar I/O port.

Here’s one

I know the comment was meant in jest. But sometimes wonderful ideas start with a joke.

Older hard wired alarm systems sometimes provide outputs that can be programmed to mirror sensor zones on output channels. This allows the alarm system to fully supervise the sensors, and also use of the sensors with some external system (like OpenHAB) without compromising the alarm. Cake and eat it,

Modern equivalent would be to use an alarm system capable of reporting activity over IP or serial interface. There are a few posts about these, various brands. So basically the alarm can be completely standalone, but still report zone events to OpenHAB,

Using a purpose built alarm takes care of standby power, callout actions, tamper detection, keypad, etc.

p.s. if you have any insurance requirement for a certified alarm, that is the way to go