Smart home – how can you control sockets and lighting remotely?

As you can see from the video, this is a device that requires the wiring (control wiring – buttons, and power wiring – light bulbs, gate, etc.) to be laid out in a STAR configuration. In other words, all the cables must converge at a single point: the electrical box with the SterBox installed. This is a viable solution provided that you are building a house and the cables have not yet been laid. In the case of an existing installation, you would need to duplicate it. Which means extra costs.
The advantage of this layout is that there’s no need for external communication (to the actuators), so the dilemmas we discussed recently no longer apply.

Looking at the price. If you wanted to build a functionally similar system yourself, you’d need a PIC for 20–30 USD, a circuit board for 100 zł, relays, transistors, other components, and an enclosure. That comes to around 400 zł in total. And you’d first have to design, assemble and program it – which requires both knowledge and time. Expanding it with additional I/O shouldn’t cost more than 20 zł per output. A project for enthusiasts. But an interesting one at that.

I don’t know the maximum number of devices the SterBox can handle (the video suggests only 16 inputs and 16 outputs, but perhaps there are more…). If there are around 40 controlled points (in a typical house with a garden, main gate, side gate, heating, etc.), you’ll need 600 + 5 × 200 = about 2,000 zł (for a SterBox with 5 expansion modules and a power supply), plus the cost of cabling (a star topology is the most expensive). It follows that 5,000 zł is the minimum.

I reckon that if you cleverly set up your own system based on a homemade PIC microcontroller and distributed intelligence, it will cost around 1–2 thousand zł (if only because you won’t need a ‘star’ configuration). And with your own microcontroller, you can support practically any number of devices. It’s simply a matter of adding an actuator and, if necessary, a bit of reprogramming. In other words, it depends on the ‘flexibility’ of your design. Expanding the system with new features won’t be too expensive.

Be smart.

Added after 21 [minutes]:

To Urgon

1. An additional CAN-UART conversion circuit is, firstly, a cost and, secondly, it complicates the printed circuit board. As few additional components as possible. Ideally, everything should be in a single chip.

2. So that brings us to the telephone modem. Maximum speed 56 kbps. Range in kilometres. It might not be a bad idea, but it does complicate the circuit by adding further components. Unless, of course, everything is implemented in software on the PIC with just a small transformer on the port… interesting. Perhaps it’s worth giving it some thought and giving it a go?

AVE...

You’ll need one op-amp, two transistors, one diode and one resistor. I was thinking of using cheap LT-44 transformers. As the secondary winding in these is split, you can select the transformation ratio using a single jumper depending on the network requirements. In that case, a single pair of twisted-pair Ethernet cable will suffice; it will be balanced and galvanically isolated. Software-wise, you can implement any communication protocol; even a simple USART with a low bit rate will suffice...

As for the star topology: if you’ve got a bag of relays, all you’ll need is a bag of 74HC595 chips, a few resistors, diodes and capacitors, and any larger PIC16F/18F microcontroller, even one costing 6 zlotys, such as the PIC16F1518. A matrix keypad can be built. The most expensive parts of the whole project will be the buttons, relays and wiring. Ten 74HC595 chips can drive up to 80 relays; if the maximum 35mA per pin isn’t enough, you can always add driver transistors. However, a star topology might only work in a single room – wiring the whole house in this way doesn’t make sense...

That’s true. You’re absolutely right. With a star-topology wiring system at home, you can control everything using the circuits you’ve described: a single controller, a 74HC595 register(s) and a relay strip. Simple and cheap.

However, the star topology isn’t ideal, particularly if you already have an existing, traditional power distribution system.
I have star topologies (signal and power) in my own home and they work well, but I can see their drawbacks, particularly when used on a large scale, and I intend to come up with something more modern.
That is why I want to create a DISTRIBUTED smart system. In my view, this involves running just 2–4 wires throughout the house (separate from the mains, of course), point-to-point: a +5V power supply and a data (bus). Actuators (relays) and event-detecting devices (buttons, thermometers, sensors, etc.) implemented on a simple PIC microcontroller can be connected to this bus. Tasks are executed in independent! actuator circuits. It is here, by listening to the bus or through polling, that the circuit reads the states of the event circuits (as well as other actuator circuits) relevant to it and decides whether to switch its ‘light’ on or off. In addition, the circuit ‘reports’ the decision it has made to the bus so that the others know what is happening. This concept is reasonably resilient to failures in parts of the installation and does not require a central unit (supervisor). This does not mean that a central unit cannot be present; indeed, it may trigger certain actions. However, in principle, it should only be used to set parameters, after which everything runs autonomously, and the supervisor unit can be removed and used elsewhere.
In addition, I’m also focusing on energy efficiency, so that the entire system can operate without mains power for up to 6 months (e.g. on 2 AA batteries). Applications: caravans and boats in winter, etc. That is why, for example, the actuators cannot be ordinary relays but must be bistable ones.

I’m also considering using relays, but that would be a last resort.
Safety is the top priority here, and relays tend to suffer from stuck contacts or other faults, such as those you’ve described.
Note that the problems described tend to relate to high currents.
With low currents, the losses are much smaller than the power consumed by the relay. )
Transistors can also lock up, i.e. remain permanently switched on, although this state lasts no longer than 10 ms in our mains supply and 8.3 ms in a 60 Hz supply – a similar effect to that seen in thyristors, and one that tends to occur with high currents.
In my case, the relays certainly won’t fit in the junction box.
The current rating must be the maximum for the entire circuit, i.e. at least 25A.
My considerations do not cover low currents, as there is no issue there.
It is difficult to see any savings in switching on something like 4,000 W, and a 25 W loss in the process is negligible.
Savings do not come from switching devices on, but from switching them off immediately when they are not needed.
Relays cannot be used to control PWM, and I want to have that option.
At high currents, even the contact resistance plays a role.
Relays are out of the question here because they would have to be large – they’d be almost like contactors.

***

When devices need to communicate with one another, the only network topology is a bus.
It doesn’t matter whether we use CAN or some RSxxx standard; LAN is not a bus.
In the case of a bus, some form of addressing is required.
RSxxx-based bus systems are highly resistant to interference and do not require the creation of complex network stacks for simple communication.
When a good protocol is developed, they are exceptionally reliable.

***

That’s exactly how I realised there’s no point in fiddling about with RS232 or RS422;
RS485 circuits draw less than 3mA; all devices are connected to the RX pin at the same time, and only switch to the TX pin when they want to transmit.
Incidentally, you cannot check line occupancy via the TX register; you need to connect an appropriately configured device – such as an analogue comparator – to the line, as I have done.
The TX register is the one you write to; you read from the RX register, but if I need to check the TX line of the slave devices, that’s the master device’s RX line.
RS485 solves this problem. ) All you need to do is handle the RX interrupt and the problem is sorted. I’ll go and have a chat with my colleagues about RS485 chips straight away; I’ll ask them to look for some for me.
If I find the time, I’ll get round to the programming later today.

***

Well, as for this communication between all the devices, I still have my own ideas in mind.
One of them is having the same software on different devices within a given functional group.
In this case, an unknown address cannot send a request, as the central unit would receive a flood of ‘softwareupdate’ requests if it sent an unknown device code.
It’s my idea, so that I can plug in a new device at any time.
All you need to do is provide the central unit with the correct parameters and that’s it; the rest happens automatically.
For me, external flash storage plays a vital role, and it’s not just out of habit from memory-hungry Windows systems. )

AVE...

Control circuit: triac + optotriac for isolation. The triacs do not need to be particularly high-rated: 10–16A will suffice for most devices. To cut off an entire branch of the circuit, a 40A triac costing just a few zlotys will do. The problem may be the whole system overheating at high power levels...

One control unit could be installed in each room, with the actuator modules connected to it locally. This controller would communicate with the main controller via a bus. Instead of polling all devices in such a network, it is better to use status change signalling. Polling could be carried out once in the event of a reset of the main controller. I’d add the option of remote management via a local network or the internet...

Why do you need PWM? It only makes sense for dimming lighting, and only for lighting based on traditional incandescent bulbs. Neither compact fluorescent lamps nor LED bulbs will work with PWM dimming...

For just a few zlotys, I can get an IGBT transistor, which performs a little better than a triac.
It allows the circuit branch to be switched off even at the peak of the sine wave.
You know, 10 ms is a bit long for devices to wait before the triac switches off.
And it certainly generates less heat.

I might need PWM, though I don’t have to use it – it’s very useful.
Imagine I suddenly need to plug a heater into a socket, then I have to go out somewhere.
All I need to do is send a text message specifying that there’s a resistive heat source in that particular socket.
After that, it all happens automatically – temperature maintenance, PWM control – there are loads of applications.
Today we have light bulbs like this; tomorrow there’ll be different ones – technology is changing.

And no 10–16A – the sockets’ current-carrying capacity must be as high as possible.
I’ve no intention of labelling the sockets with their current-carrying capacities; it’s this one’s like this and that one’s like that.

A LAN network is out of the question as a means of communication between devices.
As I’ve said, a LAN isn’t a bus; the connection is point-to-point. A bus involves multiple devices connected to a single set of wires.
Both my colleague mbi33 and I use buses in our tests; however, mbi33 has a bidirectional bus on two wires (he knows more about it himself), whilst I have a bidirectional bus on four wires with arbitration.
At the moment, I’m considering a two-wire bus with open access and prioritisation.
I want to implement collision-free communication as simply as possible; it’s not like Ethernet, where more than one device can transmit at a time.

Connecting actuators to the internet is like handing over control to strangers without any supervision. )

Do you really think a firewall is impossible to bypass?

In my setup, there’ll be internet access, but only to the central unit.
I already have a GSM modem; SMS is supported, and DTMF too – I might implement it, or I might not; it’s a bit of an old technology.
Polling is fine, but you can lose information (events) that way.
This requires queuing events in a FIFO register, along with their occurrence time.
In some cases, there might be so many events that you won’t be able to keep up with polling them, unless you poll non-stop.
And what if several devices need to poll a single device for the same event?
I don’t have that problem, because everything goes through the central unit and the central unit sends the commands.
In my system, the devices report events to the control unit, and it decides whether and how to handle them.

I’m not sure what you mean by ‘signalling a state change’.
This information has to be passed on somewhere, so either to the central unit, to a specific device, or via a broadcast mode to all devices.

AVE...

1. Are you planning to plug something with a power rating of 8 kW into a socket?
2. The idea of using PWM makes no sense, in my opinion. Specialised heating appliances and systems aren’t designed for you to tinker with a socket like that. Just install a heating system with a thermostat in every room and don’t complicate things unnecessarily.
3. Individual IGBT transistors aren’t suitable for switching alternating current. You need two transistors connected in series with their emitters touching, and a fast diode connected in parallel with each of them, in the opposite direction to the transistor’s forward direction. Furthermore, the control voltage must be 10V higher than the emitter voltage, and there’s also the requirement for galvanic isolation – without which you shouldn’t even attempt to do anything. IMHO, your control idea is more complicated than mine, and it makes no practical sense. Fuses are there to protect the installation, not smart home systems. K.I.S.S.
4. A network cable has four pairs of wires; two are used for Ethernet. So you’ve still got two left for your automation system, and that’s what I meant – not using the LAN itself. Let the LAN be used for remote home management and collecting statistical data.
5. A firewall is for protecting the internal network, not for securing the admin panel or website. If you feel a paranoid need to secure the system, use a token that generates a new hash every minute based on the time, synchronised via DCF77. Implementing this, whether in hardware or software, is not difficult.
6. A single pair of wires is sufficient for communication. You connect them using transformers to maintain galvanic isolation between the circuits. Terminate both ends of the pair with appropriately selected resistors to prevent echoes. All circuits are in listening mode. One transmits, the rest receive. During transmission, a short pulse should first be sent to inform all devices that they are to receive only; this is followed by a header indicating which device or class of devices is to listen, then the message itself and a checksum. The lower the transmission speed, the greater the range. And the packets will be no more than a dozen or so bytes in size.
8. In my opinion, it is better to divide the system into local controllers, which manage individual rooms, and a single central controller, which primarily serves a supervisory rather than a control function. There is no point in sending every request to it to switch on a light or activate a socket. Let the local controller handle this by simply reporting the occurrence of the event to the central controller. This eliminates the problem of creating a long FIFO queue to serve all interested parties.

Well, there you have your answers.

1. Who knows.

2. That’s just your opinion. I didn’t say I’d be heating it with electricity; I was just giving an example.

3. 10V isn’t much; you didn’t show off your knowledge of how much current would be needed to overcharge the capacitor. But overall, well done!

4. A LAN is a LAN and a cable is a cable; I’ll use the cable, but not the LAN.

5. A LAN is a network; a firewall protects that network. Sorry, but am I supposed to explain all this rubbish you’re writing yourself?

6. UART implements what you’ve written here at the hardware level. The range is always the same as the length of the wire. I said ‘wire’, not ‘Ethernet’. What packets? I’m only just considering packet transmission for sending large amounts of data. And it has nothing to do with the physical connection.

7. You’ve got it wrong.

8. Mechanical switches possess this sort of intelligence – local intelligence. In my system, nothing sends a request. No device can demand that another device change its operation; that’s not within its remit.

AVE...

1. I don’t know of a single load of that power rating that can be connected to a single-phase circuit...
2. OK, if you say so.
3. I found an example like this in an International Rectifier datasheet – take it up with them. Incidentally, they suggested galvanic isolation using an integrated photovoltaic cell with an LED. I didn’t even know they made such components.
4. If you’ve got the cable laid, you can do the LAN as well.
5. This network. So what if someone hacks into the LAN – the management interface is on an HTTP server, and you can disable other services and close the ports. I’m no expert in network security, but even I know that much.
6. You’re wrong here – the physical layer and its implementation do matter for the range and the number of devices connected to a single bus. How much data do you need to transmit in a home automation system like this? Surely you’re not planning to connect cameras to it?
7. I was hungry, so I ate it…
8. You can’t switch mechanical switches remotely. Breaking the entire smart system down into smaller, independent blocks – which may or may not be managed by the main controller – makes the whole system more reliable, whilst also increasing its functionality and making it easier to manage...

You’re writing better now.

But…

4. Well, not really.

5. And yet you do believe it.

6. I’m not mistaken, because how do the problems described here relate to RS485 or RS422, given the distances involved at home?

8. Of course not. You see, you need to understand the idea behind this and what I’m trying to achieve.

It’s a balancing act. When it comes to controlling 230V sockets.
The main reason for implementing such a control system is the need to switch off the load on individual sockets remotely. And please don’t tell me it can be done with relays, triacs, etc., because I already know that. The more functions I can implement whilst I’m at it, the better. ‘Remote’ here doesn’t mean from another room, but from somewhere in Poland. The involvement of a central unit is essential, and that’s not up for debate.
The actuators themselves provide control over their own operation and that of the connected loads. Collecting information on power consumption and so on.
The existence of these control elements is also non-negotiable.
This is the main reason why I’ve taken this on.
Everything else is just a bonus.
That is, lighting control, monitoring the user’s position, monitoring the concentrations of various gases, and so on.
The whole system is designed to make certain decisions autonomously, aimed at preventing any hazards from arising before I can deal with them personally. So you’ll understand that I’m not particularly interested in electrical fiddly bits.
I realise you’ve read up on OSI, hence your fascination with LAN. But LAN isn’t suitable here, as this is a point-to-point connection.

It would be good if, provided you’re willing, you could contribute your knowledge rather than pushing pointless solutions or going on about the obvious.
By ‘help’, I mean suggesting what else could be implemented without making major changes to the equipment.
I’m not talking here about, for example, additional devices that would enhance the system’s functionality.

I’ve just come across this solution:
http://www.fibaro.com/pl/system-fibaro/instalacja
A system based on Z-Wave (a wireless mesh network where every device communicates with every other device). It doesn’t require any additional power or signal cabling. The central unit manages everything. But since every device can communicate with every other device, surely the system could work without a central unit (distributed intelligence)? And I’m sure the wall switch function certainly doesn’t require a central unit.
The costs are quite high (Allegro), but the advantage is that it can be integrated almost immediately into an existing, traditional electrical installation without any modifications.
What do you think?

mbi33 wrote:

I’ve just come across this solution:
http://www.fibaro.com/pl/system-fibaro/instalacja
A system based on Z-Wave (a wireless mesh network where every device communicates with every other device). It doesn’t require any additional power or signal cabling. The central unit manages everything. But since every device can communicate with every other device, surely the system could work without a central unit (distributed intelligence)? And I’m sure the wall switch function certainly doesn’t require a central unit.
The costs are quite high (Allegro), but the advantage is that it can be integrated almost immediately into an existing, traditional electrical installation without any modifications.
What do you think?

A really cool invention – if you factor in the cost of rewiring the system and the cables needed throughout the house, plus the control unit, this device actually works out cheaper. It would be useful to hear a few opinions on this equipment.

Can someone tell me why on earth I’d want to switch off every socket in the house and control all the lighting? I simply can’t get my head round it – a socket is there to draw power from, not just to plug in the Christmas tree once a year.

I’m also setting up a home automation system, but there’s no way I’m going to bother with even a single socket; at most, the lighting will be automated in two rooms.

As for heating, I don’t see the point in getting involved with solenoid valves and a PLC to control every room either; I’ve got ordinary thermostatic heads and they work very well – after all, a house isn’t an incubator.
I’m not pointing the finger at anyone, I just can’t see what you can achieve that way – certainly not a more stable temperature, as that depends on many factors.

Ah, just 50 years ago, in a cottage, the water in a bucket would freeze overnight in winter, and the children would run barefoot across the clay floor; and today , we’ll soon die out from all this prosperity.

Example 1: The living room. A leather sofa. A floor lamp next to it. A large, arched one. Used for reading whilst sitting on the sofa. You walk into the living room and switch on just that lamp using the switch by the door. The same applies when you leave – you switch it off using the switch by the entrance. You don’t have to bend down and look for the lamp’s wall switch. Watch old American films – ‘floor’ lamps are also switched on by the door.
Secondly, when you leave the house, activating the alarm switches off any appliances plugged into the controlled sockets. That way, you won’t ‘forget’ to switch off the lights – neither the ceiling lights nor those plugged into the (controlled) sockets.
Thirdly, you can set certain ‘standing’ lights to switch on automatically, e.g. at dusk, etc. To simulate occupancy…
Example 2: There’s an iron in the drying room. You’re ironing a shirt before leaving the house. You’re in a hurry and forget to switch off the iron. Arming the alarm switches off the ‘forgotten’ lights and all the sockets (the controlled ones). Including the iron.
Example 3: The stereo, TV, etc. are connected, as above, to an alarm-controlled socket. This way, they do not consume energy in standby mode when no one is at home.
Regards
PS. The thermostats aren’t smart. They won’t turn off the heating in the bedroom and children’s rooms during the day, nor will they turn it back on before the evening. By using smart controls, you only heat the rooms you need during the day (living room, study, kitchen). At night, just the bedrooms. That means savings. And energy is getting more and more expensive!
PS. I still run barefoot across the floor because I have underfloor heating in the living room

AVE...

I’ve got another example: a storm’s coming – you press a single button and all sensitive devices are switched off. However, if lightning strikes the mains near the house, it could still knock out the whole smart home system, so a sturdy contactor at the input would be useful...

3.1.9. It is prohibited to disseminate content that is sarcastic, mocking or malicious, or that shows a lack of respect for other Users or third parties.

Next time, there will be warnings. There will be no more warnings.

SAWEK101 wrote:

Can someone please tell me why I should unplug every socket in the house and control all the lighting? I simply can’t get my head round it – a socket is there to draw power from, not just to plug in the Christmas tree once a year.

Hello
Let’s start by saying that the IB system is a really cool, expensive gadget for blokes, though women can make use of it too

I’m against integrating 230V sockets into the system; yes, a few specific points – such as a Christmas tree in front of the house – by all means – but everything else, and the example of the iron already given by my colleague mbi33 is a ticking time bomb, because whilst he might switch them off remotely, another user might switch them on when you least expect it. Unplugging all manner of electronic appliances – almost all of which have a timer, from the stereo to the fridge, microwave and washing machine – is also a waste of time; it’s easy to switch them off, but how often does it drive you up the wall when you have to reprogram that rubbish all over again? Well, perhaps once everything is connected to the internet, the issue of setting the time will be sorted.

As for lighting, my friend SAWEK101, things are a bit different here; the system can be used in a rather nice and useful way. It’s important, when preparing the installation, to include all light sources in it. By arming the alarm system with a single code, we switch off all the lights in the flat, simultaneously activate the presence simulation, switch on the outdoor lights (twilight-controlled, of course), close all the shutters, switch off any unnecessary central heating pumps, switch to economy heating mode, open the gates, and activate or deactivate other systems, e.g. the automatic watering system, heat recovery, etc.

SAWEK101 wrote:

I’m also setting up a home automation system, but there’s no way I’m going to bother with even a single socket; at most, the lighting will be automated in two rooms.

As for the heating, I don’t see the point in getting involved with solenoid valves and a PLC to control every room either; I’ve got standard thermostatic heads and they work very well – after all, a house isn’t an incubator.
I’m not pointing the finger, I just have no idea what you can achieve this way, but certainly not a more stable temperature, as that depends on many factors.

It is precisely in this area that the system offers great possibilities: day and night temperatures, adjusting heating times to the presence of household members, and a reverse function – cooling for the summer – ordinary thermostats do not offer such capabilities, and it is here, above all, that we can find comfort of use and savings.

SAWEK101 wrote:

Ah, just 50 years ago, in a cottage, the water in a bucket would freeze overnight in winter, and children would run barefoot across the clay floor; but today , we’ll soon die out from all this prosperity.

That’s life – it’s constantly evolving
Best regards

Taking this topic as a starting point, I would like to ask about sensors, detectors and various types of actuators used in smart buildings. I know there are a great many different types of these components, but most of them are used in industry and are too expensive to be used in smart buildings. What role do smart sensors play here? I’d be grateful for any recommendations for books or websites covering this topic. Many thanks in advance for your help!

It might not be all that clever, but it’s cheaper and simpler, and it’s on Android, so everything’s right at your fingertips.

Added after 49 [seconds]:

And, of course, a link to the comment

http://forum.mobione.pl/programy/nowy-program...z-cyklu-%27sterowanie-androidem%27-szczegoly/

It’s a cool idea, but spending 400 zeta on a strip is a lot; I’d rather pay a bit more and get something like this: http://sterbox.eu/index.php/sterbox. I can control it just as easily using the Android app on my phone, and I also have access via the internet when I’m not at home.

AVE...

If you know a bit about electronics and programming, you can build a controller like this yourself. In this age of microcontrollers with TCP/IP stacks and specialised chips for UART/Ethernet communication, it’s not a big problem. You could even buy a cheap, second-hand router and install some version of Linux on it; many of them have UART/TTL outputs, so all you need to do is add an external control unit with a microcontroller and write the appropriate programme for the microcontroller and for the operating system installed on the router...

I know this has little to do with intelligence, but I’m thinking of
building a sliding door on a rail, powered by a motor
the problem is that I’ve no idea how to go about it… (I mean all the
electronic control components???) because building the mechanical parts, motors and so on
isn’t a problem. I’d like there to be a button on both sides to open the door
and, for example, for the door to close after 5 seconds. (I’ve got no problem with the programming either).

The situation is exactly the same as when controlling, for example, blinds or a garage door. The motor (which moves in two directions) is controlled by two relays. These can be the simplest time-delay relays, for example DIN-rail types, or relays from the installed IB system.

What do you think, lads, of the Somfy Tahoma? You can control everything with that as well.

I think the easiest solution would be a Wi-Fi-controlled socket. For example, the Xiaomi Mi socket, which can be controlled via a smartphone. You can expand the system and control it from anywhere in the world. It costs around 100 zł and is excellent quality. I use one to control my gate in Wiązowna and my garden lighting.
Best regards

A question for those who know more about this and have been interested in the subject for longer. If I install a system like this http://instalaton.pl/16-system-fibaro, is it possible to control my home appliances remotely? So that, for example, I can switch off the TV or the iron when I’m out. Which solution or system would be best suited to this?

And how do you want to switch the TV off?
Simply cut the power via a relay module, or send a command from the server?

Mirzab wrote:

A question for those who know more about this and have been interested in the subject for longer. If I install a system like this http://instalaton.pl/16-system-fibaro, is it possible to control devices in my home remotely? So that, for example, you can switch off the TV or the iron when you leave the house. Which solution or system would be best suited to this?

The drawback of using a relay is that it completely disconnects the appliance from the mains (this does have a few advantages, but that’s another matter), so once switched off in this way, the appliance cannot be turned on, for example, using the manufacturer’s remote control – you first have to reconnect it to the mains via the relay.

Fortunately, technology is advancing, and it is possible to move away from relay-based solutions.
Currently, new LG and Samsung TV models feature an API that allows remote control of actions such as power on/off and volume/channel changes via a simple URL request. To send these commands, you can use a control centre such as Fibaro or OpenHAB, or the Android app Tasker. In this way, a switched-off TV or hi-fi system can easily be switched on using the manufacturer’s IR remote control.