Hello all - intro

Joined
7 Apr 2016
Messages
49
Reaction score
1
Country
United Kingdom
Hi all

I've just posted in the plumbing forum and noticed the home automation one so thought I'd come and say hello.

I'm really addicted to all this home automation stuff and thought I'd start with my current setup.

  • Amazon Dots in the kitchen, garage and car
  • Amazon Echo in the lounge
  • Amazon Spot in the bedroom
  • Lightwave RF sockets around the house
  • Lightwave RF plugins around the house
  • Hive Thermostat
  • Hive bulbs in various places around the house
  • Hive plugins in various places
  • Hive sensors on various doors
  • Ikea Tradfri bulbs around the house <<< These are the best thing I've bought for bulbs
  • Sonoff relays running an outside light and a couple of LED panels in the garage <<< These are the best thing I've bought for controlling other stuff where you don't want/need a physical switch.
I've also got a RaspberryPi Zero running a security camera in the porch and another in the garage. These are easily as good as, if not better than, as my friend's Arlo cameras, and cost me about £35 each to build. These are accessible from anywhere through a web interface and they upload the videos to Google Drive.

I use IFTTT for certain things as well. The cameras, for example, use motion detection and I have a webhook set up so that the lights in the garage go on if the camera detects motion. I could put a Hive sensor in there but this works just as well and hasn't cost me anything.

Lastly, and this is the bit I'm really pleased with, is the Raspberry Pi in the house that does some of the heating automation grunt work.

There's 4 people in the house and Hive don't offer proper geo-fencing. Ok, if you all have the app then it'll ask you if you want to adjust the heating but that's rubbish. I'm talking about a system that will automatically turn the heating down when the house is empty and back on when there's at least one person at home.

So, I've written some scripts in Python to do it myself. One script sniffs for Bluetooth connections to see if certain Bluetooth devices are in the house, i.e. our mobile phones. If I'm in, it sets the heating to 21 degrees as I like it quite warm. If I'm not in but anyone else is then it sets it to 18 degrees. If the house is completely empty, it waits 30 minutes to check it's not a glitch or that they haven't popped down to Tesco and then it turns the heating off.

This is all done through the (badly documented) Hive API.

So, that's me. I'll try and check in more. There's a few old, unanswered questions that I could have helped with. Don't know if there's a PM system on here but, if there is, feel free to drop me a line. This is just a hobby so I'm not touting for business or looking for payment. I work full time for the National Trust and I'm quite happy keeping the gadgets as something I do for fun.

Cheers all
Andy
 
@wwwebber Ha ha! Just read through your thread. It does make me laugh. You get one 'expert' with a massive ego that all the other sub-experts feel obliged to massage. It's never constructive. It's always about showing how clever they are rather than trying to actually help.

If it works for you then that's all it needs. If, like me, you get an enormous amount of satisfaction building something like this yourself rather than buying something off the shelf then even better. I probably wouldn't recommend doing the same if it involved anything that might cause you harm but anything else is fair game in my book.
 
In the other thread wwwebber said:
Any criminal will also have absolutely no idea what type of system I am using or the frequencies used to jam it so it doesn't particularly worry me.
Which suggests you are not aware of how easy it is to render ineffective a wireless alarm that does not use secure two way communication between sensors and panel.

I got a bit of stick from the experts but it works great for me :)
How many attempted burglaries has it prevented or reacted to ? You can only say "it works great" when it has sucessfully reacted to an attempt to break in.

As a retired electronic design and development engineer I do enjoy creating new systems. That said after 12 years spent designing communication systems involving wireless to mobile units I would never use wireless in an alarm system when all items are fixed and can be hard wired.[/quote]
 
Which suggests you are not aware of how easy it is to render ineffective a wireless alarm that does not use secure two way communication between sensors and panel.


How many attempted burglaries has it prevented or reacted to ? You can only say "it works great" when it has sucessfully reacted to an attempt to break in.

As a retired electronic design and development engineer I do enjoy creating new systems. That said after 12 years spent designing communication systems involving wireless to mobile units I would never use wireless in an alarm system when all items are fixed and can be hard wired.
[/QUOTE]

Now this is not my thread so having to respond to you here is not good news to be honest :( So, for the final time - before I had no such system - nothing !! so this is better than before - much better. Before I wasn't protected and now I am. I don't want wires going through this old house of ours and I don't want to spend too much money but I do want flexible and remote control. Sorted. End of story thank you. And apologies to OP.
 
@wwwebber Ha ha! Just read through your thread. It does make me laugh. You get one 'expert' with a massive ego that all the other sub-experts feel obliged to massage. It's never constructive. It's always about showing how clever they are rather than trying to actually help.

If it works for you then that's all it needs. If, like me, you get an enormous amount of satisfaction building something like this yourself rather than buying something off the shelf then even better. I probably wouldn't recommend doing the same if it involved anything that might cause you harm but anything else is fair game in my book.

Indeed, I'm a techie so I'm ok with all of this. So whats the next step with your setup ? where are you going next ?
 
It's going to take a very motivated and moderately knowledgeable burglar to mess with any wireless system like this. Most house break-ins are going to be done by opportunists - we're not talking about someone with a Picasso in the living room!

Yes, a nice shiny kit installed by a professional is going to be the ideal solution. Is it better to have nothing rather than something you've built yourself which, in all probability, will do the job? No, of course it isn't.

I can't afford Arlo security cameras. I've built something which does pretty much the same with Raspberry Pi Zeros for around 20% of the cost. Should I throw them away because I built them myself?

Rather than criticising with comments like "Which suggests you are not aware of how easy it is to render ineffective a wireless alarm..." why don't you offer some constructive advice on how to make this guy's system more secure?
 
Oh, I have plans for global domination! I'm an ex-techie. My degree was Electronics and Computer Science but that was a long time ago. Long enough ago that moving from an 8 bit processor to 16 bit when I was writing code in assembly language was a pant-wetting moment. I now work for the National Trust doing a lot of UX on their website.

The thing for me is learning Python at the moment. I've got my heating working exactly as I want, as per my OP. But the code is dirty with lots of duplication. That's an ongoing task but it gets better all the time.

For me, it's all about going to bed knowing a bit more than I did when I woke up. Doesn't matter how big or small the achievement, it's the buzz. Buying off the shelf doesn't float my boat, I'm afraid.

You should have a look at some of the Pi security cam stuff, if you haven't already. With the Zero and a camera, it's really cheap and you can tuck them away anywhere. The motion detection is superb and you can use it to fire off other stuff using webhooks on IFTTT.

Have a good one!
 
Your never an ex-engineer you can't unlearn stuff. Well done with your automation, I do agree we all tend to jump in with best intentions to help, but some times there are things which in practice don't work as expected.

My home now has some automation started to look after my mother, and I some times wondered what was the point manufacturers writing installation instructions as they simply did not work, every place I looked said don't fit a TRV in same area as a wall thermostat, yet it was in the end doing that which cured my problems.

Getting info is the other, I had read both Hive and Nest instructions and it seemed Nest is OpenTherm and will talk to e-TRV heads and Hive will not, however now told Hive will talk to e-TRV and looked at son's Nest and seems is not OpenTherm. I will guess down to version, but so easy to make errors.

What would be interesting is how other homes heat and cool, I had all big ideas of controlling heating so it auto turns off and on, however switch off heating over night and the temperature does not actually drop that much, also not a clue how much power is being used to heat the house, boiler output is between 6 and 28 kW but nothing to say what the output is, just it is running.

So as to geofencing my e-TRV can be set to turn heating up or down as we leave or return however the built in anti-hysteresis software means change the target temperature from 16°C to 20°C and reasonably fast up to 18.5°C, but then it all slows down, so set to raise at 6 am and it's 10 am before settled at new temperature, I cheat and ask for 22°C for 1/2 hour then 20°C and by 7 am sitting nice at 20°C but even an hour is too long for geofencing really, idea is when you get home temperature is correct, not still slowly climbing.

Now old house had a Myson fan assisted radiator, now we are talking speed, 15 minutes and room air temperature has risen 2°C the rooms heats up fast, also cools fast as the fabric of room has not heated up, only the air, it is on an old boiler that switches on/off, how it would work with modern boiler which modulates I don't know, I see modern version instead of a manual speed control has a multi speed which it auto selects to increase or decrease output, however heating is not one item, but a host of items which all need to integrate, so we are told by boiler manufacturer to use lock shield valves and TRV's but the fan assisted radiator manufacturer says not to restrict water flow, so which is correct?

And then the question, is it all saving enough to be worth it? I can see with a new build house how one can allocate room for a water heat store and have multi heat and cooling sources all supplying the same fan assisted radiators so home is held at the set temperature just like my car. However that must use more energy than allowing the home to naturally heat up during the day and cool at night, and instead of maintaining spot on temperature we really only want to limit the range, heating home to 20°C for 8 am then switching on AC to cool home to maintain 20°C may be nice for comfort, but for the pocket better to only heat to 17°C then no need for AC to switch on. But of course this is down to the potential 2 pm outside temperature, I note my e-TRV can with IFTTT link to weather reports, but are they reliable enough?
 
Interesting you should pop in today. I was just talking to someone this morning as I've done a ton of work on my webapp in the last few days and it's working much better. One of the things I've added in is to make it take the inside temperature reading from Hive every 30 minutes and store it in a database. I know it's quite warm at the moment but it was interesting yesterday to watch it throughout the day. I set the heating to 15 but the inside temperature sat at between 17.8 and 18.2 until early evening. It's also interesting to see how much (or little) it cools over night and how long it takes to get back up to the target temperature after the heating comes on in the morning. My plan is to add in something for outside temperature so I can see how that affects the rate of heat loss.

Is it saving enough to be worth it? This is the big question. There's 2 things from my perspective:

1. The satisfaction. I've said for years that my one goal is to go to bed at night knowing more than I did in the morning. Something like this means I can achieve that.

2. Financial. I've spent, probably, £200 just on smart, low-energy lightbulbs. It'll take a long time to recover that in what I save. But I'm definitely using less electricity and saving money. I've bought all my gadgets while I've got money. They're all off the balance sheet. If things get tight, that's when I'll benefit from all the little bits I've done to save on the bills. Then it'll be worth it. My direct debits have come down from £146 a month to £109. I don't know what other factors are involved but somehow I'm saving £30 to £40 a month between my gas and electric.

On heating, half the people you ask say to never turn the heating off, just use the stat. The other half say that you should turn it off. I think it depends on the thermal characteristics of the house. What we're trying to do is minimise the amount of time that the boiler is firing and I don't think there's a right or wrong solution. What I do know is that I've done both and my app plots gas units use on a graph. For me, it's cheaper to turn the boiler off when the house is empty than it is to, say, set the heating to 15 degrees.

And there are so many other characteristics to take into account. Our house has one wall which is very exposed. It's 30 feet long and 2 storeys tall. That's a huge surface. If the wind is hitting that then it's a massive heatsink. Computers have metal blocks to dissipate heat on some components. If they want to dissipate it faster then they add a fan. The same principle applies to a house. Big surface, windy day, and the heat is going to dissipate faster than on a day where the ambient temp is the same but there's no wind. I'm not convinced there's a way to automate things 100%. Sometimes it's going to take a wet finger in the air and common sense.
 
Hi all

I've also got a RaspberryPi Zero running a security camera in the porch and another in the garage. These are easily as good as, if not better than, as my friend's Arlo cameras, and cost me about £35 each to build. These are accessible from anywhere through a web interface and they upload the videos to Google Drive.

I use IFTTT for certain things as well. The cameras, for example, use motion detection and I have a webhook set up so that the lights in the garage go on if the camera detects motion. I could put a Hive sensor in there but this works just as well and hasn't cost me anything.

I need something like this. We want CCTV but I don't have a spare £300 to buy Hikvision kit. Also, makes sense to save to the cloud rather than on a box that can be nicked.

fancy posting an easy to follow tutorial?
 
Ok, it's not going to be as easy as sticking your hand in your pocket and buying an Arlo but the outcome will probably be better and you'll have something you can replicate. The software I'm using now is here:

https://elinux.org/RPi-Cam-Web-Interface

This has instructions on installation.

When you set it up, it'll ask you what port you want to use. Set it to something high and odd. I'd go for something in the range or 8500 to 8800. It won't conflict and it's less likely to be guessed.

I don't know if you have much experience of setting up a Pi but I'm assuming that you can do this. I'd do it with the lite installation of Raspbian rather than the one with the desktop, just because I'm happy with the command line and I don't want the bloat of a desktop OS. You can get all that here:

https://www.raspberrypi.org/downloads/raspbian/

Installation of the OS on the Pi here:

https://www.raspberrypi.org/documentation/installation/installing-images/README.md

The first time, it will take you an hour or 2.

I've done this a few times and I reckon if a box arrived from Amazon with all the bits right now, I could have a camera working within 30 to 45 minutes.

So, that should get you to a point where you can view your cam remotely if you're connected to your network and have motion detection to save videos onto the SD card. If you want them on the cloud, the best bet would be to setup a Google Drive account and install rclone:

https://github.com/pageauc/rclone4pi/wiki

There's a section on there about automating it with crontab. I have mine running every 5 minutes. This means there is still a danger that your cam will get nicked with the evidence still on it. But it's a deterrent and if you put the camera out of the way then it probably won't get seen. I use the Pi Zero with the proper Pi Zero case. I also bought a couple of lenses from Amazon... the sort that clip onto your mobile phone for wide angle, etc. I've super-glued one of these onto the front of the Pi Zero case to give a better wide angle pic on the cam. It works amazingly well!

The last bit is if you want to connect to your cam from anywhere. You'll need to set up port forwarding on your router. I'd connect the Pi to the WiFi, find out what the IP address is, reserve it on your router and then set up portforwarding on the router to point that port on the outside facing IP address to the IP address of the Pi on your home network.

This might sound hugely complicated if you've not messed with Pis before. But it's really, really easy. If you decide to give it a go and get stuck, shout here and I'll do my best to help. If I can't, there's a forum for the software on the Raspberry Pi forum and you'll get answers there Quite quickly.

Good luck!

Andy
 
Back
Top