Smart Home: An Example of Development and Implementation. Part 2

In the second part of our series on the smart home I will share my experience with developing a smart home system for a two-room apartment. In the first article I talked about my approach towards designing and developing a smart home system, and in the third one I will look at the solution that has been operating for 4 years already.


Once you have clarified all your needs, you can focus on developing a smart home architecture. At this stage, it’s important to account for the size of the premises to be controlled by the system. First of all, you should decide on the topology of the power supply and control lines. An ideal variant is the “star” topology where their own power lines go to each consumer and all the control electronics is located in one panel near the traditional power supply board on which switches are installed.

A less expensive option is to use the power distribution network called “global bus” and to install a control relay in the chain immediately before that particular appliance. A drawback of this approach is that it makes the relay maintenance more difficult as it may then be covered by design elements such as wallpapers or fake walls. Today there are relays that can be installed directly in electric junction boxes together with other electric equipment

The next important thing that determines a smart home architecture are the control links. There are two types of control links:

  • Separate communication lines to each relay, or a group of relays controlled through one communication line. The advantage is that you have high reliability but it comes with higher costs on additional communication lines.
  • Radio channel control. Each relay is controlled by commands transmitted through a radio channel. The benefits of this method include simple installation and the absence of additional control lines. A minus is that the radio signal may die out and then a command cannot reach the relays. An example of such relays is the Noolite system (, which is used by me.

For economic considerations, I chose the radio channel for my smart home. To make the system more flexible, a separate power supply line was laid to each “consumer.”

The most important thing for smart home architecture is to select a control scheme — decentralized or centralized. Let’s discuss each one in more detail.

In the decentralized scheme, all appliances communicate with each other, there is no single decision making center. Here “each-to-each” communication is realized. Actually, it’s not that different from a standard light control scheme, just that switches and relays have got more “intelligent”, being able. for example, to turn off the light by timeout or turn on several lights controlled by different relays with one switch. Such a scheme allows for a smart home controller which receives notifications about the state of light sources and can (according to its logic) transmit control commands to relays.

The above scheme shows that a command from the Button goes directly to relays connected to it, yet this same command is also passed to the controller which is just informed that the Button was pressed. At the same time, the relay informs the controller about changing its state. This scheme is absolutely reliable, as in the case of the controller’s failure the lights will operate even though some additional functions won’t be available. On the other hand, such a scheme is less flexible, as it would be more difficult to implement changing the logic of operation from one switch with different relays depending on the time of the day, for example. A centralized scheme would be more suitable for that.

In this scheme, all the commands go to the smart home controller and all control is done by this controller. The main disadvantage of this scheme is a single point failure. Once the controller fails, the entire system becomes inoperable, yet an important advantage of such a solution is a high flexibility in managing various devices and the option to change the logic of operations depending on individual preferences. Given that the probability of controller failure is very low (during four years of the system operation there was only one serious failure due to the memory card), I selected the centralized scheme for my smart home architecture.

In the next article we will look at the solution I have implemented successfully for the past 4 years.

Pavel Tsytovich
.NET & C++ Consultant

Originally published at