Smart products are around us, and we live in an era where entire houses are becoming smart! What are smart homes? Smart Homes are enhanced houses, designed to assist residents in their day-to-day activities. Smart Houses are becoming more and more common in the world, their numbers increasing every day. What once took Bill Gates 12 years to make 2 decades ago, you can now make in probably 3-4 months. And in this article, we shall cover exclusively about such homes, Smart Home products and technologies involved, and the guidelines on how one can set up Smart Homes.
What is Smart Home Technology?
Before we begin, it is important to define the word ‘Smart Home Technology’. Home Automation, or more commonly referred to as ‘Smart Home Technology’ is the use of technology to automate your home, with the help of Internet of Things (IoT). Smart Home devices come in a variety of products, each with their own feature, and they can automate almost every aspect of your day-to-day household activities. The first example of this in the world was the development of the Thermostat, after which many more products soon followed suite. The two major types of devices used in Smart Houses are:
- Active devices, such as switches and control panels, which the homeowner can manipulate to some extent
- Passive devices, such as sensors and receivers, of which the owner has no direct contact
How do these technologies work and what are their types?
These technologies use a variety of sensors to capture information (such as IR, UV motion, auditory, cameras) and internet technologies (IoT, cloud communication, SQL databases) to make use of the data captured. They vary from country, company, and type of automation and there is no universally accepted protocol. The first industry standards for Home automation were the EIB (European Installation Bus) fieldbus and X10; and these go all the way back to the early 2000s. But today, Zigbee, Z-Wave and Insteon are also leading Home Automation protocols. These new standards rely heavily on technologies such as mesh networking.
There are two types of home automation:
- Central control: From a single control panel, you can control almost any aspect of your house, and change their properties. This panel is directly integrated with the wiring of the house and hence there is a physical connection to the main power supply.
- Pros: These systems have a one-for-all control system, and hence it is very convenient for us to choose the right settings for the house. Usually, these home automation products are very upscale and can be found in high-end luxury apartments.
- Cons: These are expensive and do require professional help for installation. Another drawback is that you can only include products in the network which are compatible with the system. Moreover, the removal and inclusion of a device often require cumbersome settings changes every time a change of device occurs.
- App based: These are becoming more common, and use Cloud technology to communicate with the internet, and IoT to communicate with other devices. Many Wi-Fi technologies such as mesh networks uses a range of extenders to allow for smooth connections across the house. With the use of such Wi-Fi technologies, many devices connect to the cloud. This way you can control them anywhere from either an AI backed master device (such as an Echo speaker from Amazon which uses Alexa) or directly from smartphones and PCs. Some smart home devices use Bluetooth technologies as well. These are becoming more and more commonly seen in households.
The biggest advantage is the accessibility. Anyone can easily access and install these devices through their personal devices. The other advantage is inclusiveness, as many new smart home devices in the market have compatibility with today’s popular AI software programs (an example being ‘compatible with Alexa’ or ‘compatible with Siri’)
The problem is that many devices can have their own separate apps to use, and this can become very tiring for the end user to use every different time. There is a risk of privacy breach and malware attacks if your devices are not protected by some sort of firewall and internet security. This is another disadvantage.
These are the two broad categories, but these days many automation companies are trying to integrate both these technologies (have a control panel whilst not keeping it exclusive) in many houses so that there is a seamless home automation experience for the end user.
What are the available products in the market today?
There are many available devices in today’s market, and these products can be classified generally by their use in the respective areas of the house. Most of these products are items which you probably know and use in your daily lives.
- Smart Home Basics (all the products for your living rooms and passageways)
- Smart Speakers
- Plugs (can automate non-automated products once connected to them)
- Light Automation
- Video Doorbells
- Robot Vacuums
- Smart Kitchens (Every item in this list is smart)
- Dishes (plates that can monitor your calories and items and suggest efficient meal plans)
- Slow Cookers
- Smart Coffee Pots
- Smart Trash Cans
- Dash Buttons (by Amazon, order necessary daily use products by simply pressing a button. For example, this allows you to place a ‘Tide’ button near your washing machine and you can order detergent every time it’s depleted)
- Smart Bathrooms:
- Smart Toilets
- Enhanced Vanities (examples include automated mirrors that can show you and tell you useful information on the side, drawers that know when you have woken up and provide you with pre-heated towels, etc.)
- Smart Showers
- Privacy Windows (which turn opaque whenever it senses that you have entered the showering area)
- Smart Nurseries
- Video monitors
- Smart Rocking (to keep your child engaged while you are out)
- Health Monitors (live updates about your child’s health)
- Changing Tables
- Smart Mobiles (Can sense your child’s pattern and use AI suggested operations to put them to sleep)
- Smart Pet Care
- Video Monitoring
- Self-Cleaning Litter Boxes
- Pet Doors
- More Smart Options
- Smart Fans
- Automated Sprinklers
- Smart Pools
- Smart Beds
These are just some of the available products on the market, which we have tried to classify and summarize in a nutshell. The main concept of all these devices is the same – that they connect to a cloud network mainframe and can be accessed and controlled either by a master AI or by human intervention.
Smart Homes need to be set up and configured in a certain way. And yes, there are certain guidelines to be followed while doing so. The following sections will help you with setting up your very own smart home with a set of useful guidelines and tips.
Factors to consider for building the Smart House
There are tonnes of features available in the market today, but not all of them may be necessary for us. Hence, we must carefully consider a large number of factors, such as:
- Why are we building the Smart Home?
- How can the system keep checking up on itself every now and then so that it can detect faults before they happen?
- What is our budget?
- What services do we require (E.g. Automation, Assistance, etc.)
- How can algorithms be created so as to recognize daily activities of living of residents?
- Can appropriate moments be identified for either aiding or providing guidance?
- How can a system automatically adapt to the resident’s daily habits?
- What can we do to make the system become more energy efficient?
- How can the network and technology be built so that there is a robust system in place to support failure?
- Even though there is a considerable amount of automation, what can we do to ensure sufficient manual control?
- How can the system reduce data processing complexity so as to have more efficient computing capabilities
We shall now break down the following sub sections into guidelines for the Hardware components, Software used, security measures that must be taken, and finally provide a flowchart with an explanation of the procedures needed.
One must carefully select the hardware for smart homes. The problem is that there is no clear guidelines for how to choose them and furthermore, scant existing ones focus on new homes (an ideal scenario) rather than existing homes (which is most of the cases).
1. How to choose sensors?
We need to exploit sensor technology and select them carefully for our usage. There are 2 POVs to look at – user side and system side. On the user side, we must make sure that the cost is low priced and affordable for the user, both on the run and installation costs. We also should take care about the robustness and efficiency while selection. Also to consider is a thorough construction.
On the system side, the installation should be easy. The sensors should be precise and not too complex in the information they transmit (if it is complex, then the computer will take longer times to process). The next two parts will be covered in the following sub sections.
2. Energy Efficiency
Energy efficiency is a system requirement that is particularly important to the end user. Many companies are focusing on their products to be very environmentally friendly and energy efficient in terms of operation and disposability. The resident cares about the bills he pays and might want to reduce his footprint. So the use of disposable batteries in sensors should be avoided as much as possible.
3. Perception of the resident on sensors
Studies have shown that residents are often aware of and perceive their surrounding habitats. By this, we mean that sensors should not invade into the privacy of the residents, and one must carefully select the sensors and effectors to minimize the negative impacts of invasiveness. They must be hidden from the view of the resident in the house.
4. Description of the types of Sensors
Security alarm and CCTV surveillance system vector illustration. Isolated set of digital video cameras with infrared sensors for home or property safety monitoring against burglary theft intruders. At the end of this section, we’ve included a table that will help you compare with ease on how to choose the sensors.
5. Video Cameras and Microphones
Video cameras are great, in that they can offer the functionalities of many different sensors at once, hence being greatly expressive. They are very affordably priced and can be constantly employed. But the downside is the feeling of privacy invasion, and on top of that the data complexity is large (leading to the use of elaborate AI algorithms) due to image analysis in different lighting conditions.
Microphones are similar, in that from auditory data, a device can process information. But a high decibel background will render it less effective, and there is always the fear of invasion of privacy.
6. Smart Power Analyzer
These analyse the power reading of electrical outlet throughout the house. Low cost models with a greatly developed algorithm can do the trick, with simple installation and robust operations. The downside is that it doesn’t cost less than $1000 USD; it must be connected to the electrical box of the entire house and also, each electrical device must be manually labelled (since they have a unique signature).
7. Radio-Frequency Identification (RFID)
RFID is a recent technology which is generally expensive (due to software/firmware on the collection module) but the supplementary tags or antennas are cheap. There are 2 types: Passive and Active. Active can be used to track moving users around the house but are expensive and require batteries. Plus they are slightly invasive. Passive on the other hand are cheap and inexpensive, but the drawback is that it is a little imprecise. The simplest way of installation is through proximity-based localization.
8. Ultrasonic Sensors
There isn’t much research on them but are supposed to be able to be a better camera version, and also give a clear 3-D image of the environment. They are less invasive due to their small space.
Figure 1 Comparison of the most common sensors from A (Best) to E (Worst)
9. Communication Technologies
X10, the older system, used a Power Line Carrier (PLC) standard where the existing home wiring is used to operate. Disadvantage here is that not all components are compatible with the PLC. Hence we should use Wireless mesh technologies, Bluetooth, RFID, UWB (Wireless USB) and Zigbee standards.
10. Centralized or Decentralized processing
Basically in centralized, the components are dumb and report back to a single server while in decentralized, the components interact with each other and take collective decisions based on AI Algorithms. More efforts and research is being put into decentralized systems. But currently researchers are trying to naturally progress from centralized to decentralized for better understandings. Furthermore, It is easier to make decentralization on the software side.
11. Choosing the right effectors
Collecting information from sensors on residential activities in smart homes is important, but it wouldn’t be complete if there are no methods to react or aid the resident. On the basis of the data provided, useful prompts are provided.
12. Categories of Prompts
There are 3 categories of prompts: auditory (E.g. verbal instructions, sound alerts, musical – transmitted through a speaker system), visual (E.g. colours, shapes – transmitted via screens or projectors, LED lights) and video (E.g. video of a person performing a task – through TV or projector screens with speaker systems).
13. Guidelines to choose prompts
Prompts should be chosen based on the needs of the residents. They should take care of any medical challenges the person may face. If someone is colour-blind then the installation of LED lights is not acceptable. If a person is impervious to sensory prompts, then one must find other ways to prompt them so as to aid them better. Also, if one is using video prompts as their effector, then the installer should take care that sound is paired with video.
Software applications should be included to provide a layer of abstraction to work with the smart home infrastructure. The reason is that this step allows the installer to remove the need to redo communication wirings with all the different components.
1. Addressing the heterogeneity
Software side of a smart home has the important role of creating a uniformity amongst various heterogenic technologies in the house. Traditionally, this problem is solved by using a middleware, which process input data from various sources and converts them into uniform outputs. This however isn’t the best choice always as AI execution depends on data complexity and volume, both of which is small in the smart home case, and also that the future changes can be a tedious task. A database can be an alternative for it supports multiple languages for creation
2. Calculation complexity
An AI software must be fast and responsive to the residents of the house. Generally, the accepted times for such a software is only around 200-300ms (less than a second!) The processing must be done instantaneously on a human timescale. If say the calculation complexity is greater than quadratic, then even adding more processing power won’t be enough. Plus it is desirable to minimize data space occupied by the AI.
3. Designing for the resident
A smart home must encourage the residents to take decisions themselves, and it is important that they feel in utmost control. AI programs can age gracefully and learn on its own but an algorithm can be dumb in a few years, which one should consider.
4. Energy management
A resident won’t be eager to install a smart home setup immediately, and one concern to that is energy management. Smart homes needs intelligent AI technologies, targeted towards efficient energy operations. For example, it must predict temperature needs and based on collected data, adjust the ACs and thermostats. If someone is sleeping, it should get the best possible temperature. The point is, it’s important to prepare foreground for rapid AI prototyping that allows abstraction and simultaneous control of Multiple Devices.
In summary, we must make sure that we follow a very systematic and procedural approach to the entire process so that we can achieve the most efficient and targeted approach. The following flowchart summarizes in 5 key steps, divided into 3 separate stages. We have to take care of the needs of the system, the houses, and the individuals, and provide them with a stable cost vs efficiency outcome that will benefit them in the long run. The AI must have great decision taking capabilities which uses efficient logical designs that not only saves on memory and computing but also on energy consumption and data storage. Lastly, one must ensure that any decision taken, whether hardware or software, is taken with care as to provide as little maintenance as possible.
Figure 2 The methodological process of iteration of a smart home.