2.4. Smart-Home Controllers and Smart-Home Personal Assistant Systems

Another type of technology used in a smart home environment are smart-home controllers and smart-home personal assistant systems. This type of technology is able to connect to various other smart devices and through them, to the IoT sensors, as well as to smartphones and tablets.

The role of the smart-home controller is to solve the problem of technological fragmentation which arises in the implementation of the ecosystem of smart homes. Different IoT devices use different communication protocols, which can be confusing for smart home users. One solution would be to create a universal protocol that would be used by all devices in a smart home environment. However, such an approach seems unrealistic, as devices are highly heterogeneous with different functionality, and it is difficult to design an all-in-one protocol for each device from different manufacturers. Another, simpler solution is to implement a gateway that supports many protocols, which act as a central controller for all smart devices in the home. Such a gateway is a smart-home controller. However, even in this case, manufacturer’s list compatibility, as there is a huge number of different IoT devices on the market and it is not easy to support each of them [32]. In the case study, the smart-home controller Athom Homey is used, and it is also possible in this category to include also the above-mentioned Fibaro system.

On the other hand, the goal of smart-home personal assistant systems is to eliminate the process of typing. So that the user does not have to type a query into the Internet browser, the dictation method is used here, where the user enters the query as a voice request. Based on speech recognition, the Smart-home personal assistant converts the request into text, performs the required action, and announces the result again by voice. The systems use artificial intelligence techniques such as speech recognition, natural language processing, dialogue systems and speech synthesis [33]. These include, for example, the Amazon Alexa, Google Assistant, Apple Siri, or Microsoft Cortana.

With the advent of smart-home personal assistant systems, users have acquired a number of new intelligent services. They aim to provide a wide range of services by consistently collecting various kinds of information such as personal profiles, location or time. While the IoT concept dates back to the 20th century, the connection to personal assistant systems was not available until about 2014 with the Amazon Echo release and the integration of IoT devices into the agent’s affordances. The interconnection of personal assistant systems and the IoT field bonds the area of virtual agents with their interaction with physical objects and users, together with networked physical devices, including sensors [34]. The personal assistant system is intended to serve as a user interface for connecting to a smart-home and to pick up and interpret voice requests to utilize other devices and services, either using scheduled actions or voice orders. Managing smart home devices through a smart-home personal assistant system obviously requires connecting the smart-home personal assistant system to the device.

The architecture of smart-home personal assistant systems for communication with IoT devices is designed in such a way that communication does not take place directly with IoT devices, but requests to go through the cloud. An example of a smart-home personal assistant architecture is shown for the Amazon Alexa ecosystem in Figure 8. A smart speaker device for voice control such as Amazon Echo Dot is required to record a voice request. The key component is the cloud, in this case the Amazon cloud platform, which includes Alexa Voice Service (AVS) as well as other cloud services such as authentication, data management, and logging. Access to cloud services requires a personal device to run companion Alexa applications, such as the Amazon Alexa application for Fire OS, iOS, or Android, referred to as companion clients. Users can also use web browsers to access the cloud. Alexa ecosystem can be extended by connecting compatible IoT devices, their cloud services and third-party applications for various other services [35]. This approach has an impact on the reaction to a possible DDoS attack on the smart home environment as well as on the resistance of IoT sensors to DDOS attacks.

Amazon Alexa ecosystem.

The case study did not use the voice recognition of these devices, but the control of the respective IoT sensors via mobile applications smart-home personal assistants.

Homey is a smart-home voice-controlled device created by Netherlands-based startup Athom, Figure 9. Homey supports a variety of communication techniques for home automation, including Wi-Fi, Infrared, Zigbee, Z-Wave, Bluetooth, and 433 MHz and can communicate with a bunch of differently configured gadgets at once. Homey is a voice-controlled device, i.e., it is possible to send voice commands directly to Homey and so to control devices added to Homey ecosystem.

Smart-home controller Athom Homey Pro 2.0.

It provides its own mobile application for managing smart devices in the home, as well as for displaying data from IoT sensors. The service consists of multiple wireless technologies enabling connections to wireless devices, and congregates these on a single platform which connects to the Internet. It uses local or cloud secure connection. Homey Cloud Backups automatically creates a backup of Homey every night [36].

Athom Homey Pro 2.0 was used for the experiments performed, and the control of connected Philips Hue bulb and Fibaro Wall Plug was handled by a Homey mobile app provided by the Athom.

Amazon’s Alexa is an intelligent personal assistant developed for use with the Amazon Echo. The Amazon Echo family of smart devices with a microphone and speaker connects to the intelligent cloud-based voice service, Alexa Voice Service (AVS). With Alexa as a voice-activated personal assistant, Echo is able to do a variety of things, such as manage to-do lists, play music, set alarms, place orders, search for information, and control other intelligent devices. Another great aspect is that Echo can also control the third-party hardware and software attached to it. It can control IoT devices such as smart-bulbs or play music from several streaming services.

The user can also define the so-called Skills. Skills provide a new channel for new content and services. Skills can be triggered by specific voice commands and allow customers to use their voices to perform daily tasks such as checking messages, listening to music, playing games, and more.

The Echo device is set up either using the mobile application for iOS or Android or via the website https://alexa.amazon.com. Echo requires a connection to a local Wi-Fi network and requires a constant connection to the Amazon cloud, as all voice requests are processed in a cloud environment. Echo and companion apps do not communicate with each other, but both are connected to the Amazon cloud, which acts as a communication proxy.

The Amazon Alexa ecosystem makes it available for users to place multiple devices in their homes into a comprehensive smart home system. For management purposes, Amazon provides functions for creating different groups of devices, assigning devices to specific rooms, assigning Echo devices to connected devices, and more [37,38].

Amazon Echo Dot 3 was used for the experiments performed, Figure 10, and the control of connected Philips Hue bulb and Fibaro Wall Plug was handled by an Amazon Alexa mobile app provided by the AMZN Mobile LLC.

Echo Dot (3rd Gen)—Smart speaker with Alexa.

Google Assistant is a smart personal assistant developed for use with Google Home. Google Home as a smart home centering product was launched by Google Enterprise in 2016.

Google Home functionality consists primarily of the following features: (i) information retrieval—Google Home allows users to perform voice searches and life reminders; (ii) multimedia control—Google Home can act as a home wireless multimedia center with an integrated wireless speaker device; (iii) environmental control—Google Home allows to centrally control smart devices in a smart home system.

Among them, the main function is the voice assistance of Google Assistant, where the system responds based on the user’s voice requests. With Google Assistant and Google Home, the user can manage a large number of services and smart devices including sensors and actuators. Google Assistant is a service that uses artificial intelligence, machine learning, speech recognition, and collected data. The main processing is not performed on the device, but on the cloud [39,40]. Google Home is operational with wake word, “Ok Google” or “Hey Google.” Parallel to the Alexa Skills, in the Google environment there are the Google Actions.

Google Home is primarily a cloud-based service and requires a connection to a local Wi-Fi network. However, when the main router is offline or if it loses Wi-Fi connection, Google Home starts to broadcast a local Wi-Fi beacon for set and, consequently, the Google Home goes into set up mode where its own unsecured Wi-Fi is established.

Google Home EU was used for the experiments performed, Figure 11, and the control of connected Philips Hue bulb and Fibaro Wall Plug was handled by a Home mobile app provided by the Google.

Google Home EU.