Advice designing an alarm/home automation system

Using Arduino Project Guidance
1

Hi,
I have been thinking about making a home automation system and an alarm system for several years in the back of my mind, and im starting to think about actually doing it.
I would appreciate any advice from people regarding their experience with similar things, and peoples advice on off the shelf modules that can be affordbaly integrated.

The general plan would be to have a central hub and a pro mini with some kind of radio modem and sensors attached to it to act as an alarm sensor for each area needed.

The requirements I have are specific on the home automation system due to a specific scenario at my home, IE I cannot easily draw more than 2KW power from the grid at once and the damn smart meter trips and I live in an old house that he elec board doesnt want to upgrade, even though it recently was on a higher tarrif (long story)

to deal with this, id like to make something I can add inside a normal socket extension lead, that senses current , and can communicate with the network, then cut the power to the device via a relay when needed. Then I can use the Hub to calc the total load on the system, and prioritise certain devices over others and power off / on as needed.

So I would like to achieve the following:
central alarm "hub" with a screen and simple 4 or 5 button keypad.
alarm sensors. (pro mini with sensors attached) comms via a radio link
automation devices - current sensing, radio modem, relay.
Automation - water ball valve actuators.

if anyone can help find a decent affordable wireless link, thats easy to interface that would be a ggreat start, I will lkely get some ball valve actuators on aliexpress. im stumped for AC current sensing boards.

for alarm sensors, I plan to use reed switches on doors and windows, and vibration detection on doors and roof. I wilL buy an off the shelf siren and hack it to control it.

finally, I may well at some point want to try to add a meshtastic node to the "hub" so that I can have the alarm ping other devices at a longer range in the event its triggered.

any help finding the components to begin with will be greatly appreciated, Wireless coverage between nodes would likely be up to 50m with a couple of block walls between them, so I may implement some message forwarding between nodes if I find they cant all reach the hub seperately.

thanks in advance, hopefully someone has some good advice

2

Use a wired alarm system. RF jammers are readily available. Only use wireless as a last resort.

3

having to program two microcontrollers with associated communications protocol makes the overall task an order of magnitude harder
I would recommend rather then using a microcontroller with a seperate WiFi module to use one with a builtin WiFi capability, e.g. ESP32 or ESP8266 - both powerful microctntrollers with plenty of IO capability

to measure current load I would recommend using a none-invasive clamp such as sct-013-000-ct-sensor

Edit: do you have WiFi throughout the property? if not look at ESP-MESH

4

My system uses CAN for node to node. I tried wireless and it has problems and is not reliable or constraint.

5

I have just done a project with two pro minis communicating over i2c and wrote my own hashed up protocol to take care of about 20 variables and not be able to confuse them if data is lost etc..... im confident the same approach can be used over a wireless link.

An there is no WIFI over the property. Although, the "shed" that will require the alarm, will have a spare cat5 run to it via a buried pipe. perhaps two spares along with a spare 2mm2 section of cable so i can run a UPS in the house to power certain items in the shed, and house power for the rest. (I want the UPS in the house as it will interfere with my ham radio if in the shed)

6

I could consider such a system, but over 50m wire runs i dont know how reliable CAN is.?

im only looking for say one data burst from sensors every 20 seconds , perhaps 30 seconds under non triggered circumstances, and a heartbeat of a byte of data every 5 seconds or so to make sure the nodes are alive.

as long as some data is received for sensors under non triggered moments, I can discard any corrupt data and wait for hte next burst, its not going to be super critical.

7

Im not expecting there to be any tech savvy theives trying their luck. and if they are tech savvy, well they will have to be quite savvy to crack my half baked protocols. lol .. ..

the basic plan is to poll for node heartbeats every few seconds, so just a blip of data. I may code a lookup table for the correct reply into the heartbeat..... this will mean that if they dont have the lookup table they would be very lucky to have the right data to spoof with,,,, and if the heartbeat isnt received, well depending on the circustances and state of the alarm, that might be enough to trigger it...... jamming can always be used to trigger the alarm, especially when I live in a remote area with almost ZERO noise floor.

imagine the node waiting for a received ping, decoding the received ping and sending back the correct reply, every 10 seconds or so, who knows. If no reply is received, the HUB can ping another node on the network, if that also doesnt reply, I could consider the possibliity of a JAM being underway, increase ping frequency and ping all nodes, if they dont reply, assume its being jammed and activate the first stage of the alarm, IE external lighting down by the shed, and a low key alarm in the house to let me know to check the shed, give it another moment or two and if i dont acknowledge the low key alarm in the house, ramp it up so the neighbours can hear it.

How ever I end up programming it, the logic will be sound, and set to avoid false alarms and annoyance of the neighbours, but also without compromising security. If im going to bother doing it I will do it slyly.

8

this is the sort of thing im looking for yes. thank you

9

Whatever you built, I'd recommend building something compatible with a standard home automation protocol. There are several but Z-Wave and Zigbee are the most popular.

All of the big players (Google, Amazon, Apple) are standardizing on Zigbee, but with right hub most of their stuff will work with other systems.

Something like this dimmer switch that fits into a standard switch box, works remotely or manually, and doesn't require re-wiring, is very difficult to build and it would probably cost more to build than to buy.

And you are using a standard protocol, you can expand your system and add something (like a controlled-light) without designing and building something from scratch.

I didn't build any of my system myself, but I've got a Z-Wave system, and I chose Z-Wave because there was more "stuff" available at the time. And I have a Hubatat hub. The hub is programmable and it runs stand-alone and turns my outside lights on a sunset, and off at sunrise, and some other events are programmed. I can also access the hub to control everything via Wi-Fi, but my system isn't accessible remotely over the Internet or cell phone network.

10

As an alarm installer.... not exactly true. Yes, a home-made alarm is probably easy to crack, but the wireless systems we install (not cheapy Argos/B&Q ones) are pretty secure.
Never had one hacked in 15+ years and my wireless system has been bulletproof for 10+ years.

11

i think wireless is secure as long as its hard to spoof and can detect jamming attempts. of course if sensor batteries run out then it becomes less effective :slight_smile:

12

Depending on baud and configuration it is specified over 1000M. Consider it is one of the primary communications busses in vehicles. It is used in automation and thousands of other applications. It is very robust and fault tolerent. Nice part it is multi-master. Here are two links to check out. https://www.autopi.io/blog/can-bus-explained/
CAN bus - Wikipedia

13

The lithium batteries in my PIR's and door sensors have lasted 6 yrs so far.... that's prettty damn good.
Checked their status on the panel a while back and all still doing fine

14

that is pretty darn good, I must admit that if i could interface with commercially available sensors like that then it would be better. they clearly have a good lowe power mode.

Ive not played with low power modes in the arduino yet

15

I don't think you could hack commercial sensors too easily. Some things are just better bought as a kit!

I also like that I have a wireless microwave sensor in the rear of my van, and when I get home at night, it is within range of the system and I can part set just my workshop and the van.

16

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.