The year 2018 saw the final breakthrough of voice assistants integrated in home speakers offered by some major international companies. From little more than gadgets to recollect a piece of music or listen to weather forecasts, voice assistants are becoming increasingly powerful instruments, with reflections on many activities of daily life. Now these devices are also proposed as an easy and intuitive control interface for the entire smart home, allowing the most immediate control you can think of, simply by saying the task to be done.
Home speakers and voice assistants: just a fashion of the moment?
Home speakers with voice assistants should be recognized first of all the extreme simplification in the relationship between the physical world and the digital world, including, of course, the home automation system of a smart home. The operation of these devices is now well known: a home speaker always listens and is activated only when a specific phrase is pronounced by the manufacturer. Who has never tried to ask for anything by saying "Alexa" in front of Amazon Echo or "Ok Google" for Google? But behind the apparent simplicity and affordable price of these devices, lies in fact a rather sophisticated technology that brings into the homes of each of us the future in the form of artificial intelligence and machine learning. These are two fundamental features: while the first allows voice recognition through special algorithms, the second enables the devices to learn and improve over time their skills. It is perhaps surprising that voice control is not a real novelty and had its predecessors twenty years ago. At the beginning of the millennium there were already technical solutions that allowed to control a home automation system by pronouncing precise sentences. However, these devices were rather expensive, challenging to configure and put into service, designed primarily to facilitate the lives of disabled and elderly people who were not self-sufficient.
In 2019 the scenario changes
The recent entry into the field in Europe of operators OTT (Over-The-Top), the obvious reference is in particular to the companies Amazon, Apple and Google, opens other perspectives. With a few dozen euros you can get an intelligent loudspeaker with an integrated voice assistant. According to IDC (International Data Corporation) research, in 2018 16 million intelligent speakers with integrated voice assistants were sold, of which 7.5 million in the fourth quarter alone, thanks to the drive represented by Christmas purchases. And the IDC itself predicts that 48 million home speakers will be sold in Europe in 2023. Compared to the pioneering phase, the purchase of these devices is no longer determined by factors such as curiosity or impulse, but by the growing presence in homes of connected objects and home automation solutions.
The interface to KNX
By means of a gateway, the most common home speakers can also be interfaced to the KNX home automation system of a smart home. In addition to voice commands, ekinex can also be controlled with the Apps offered by the OTT. The interface to the home automation system is rather easy. The gateway configuration tool allows the import of the project carried out with ETS, an operation typically carried out by the system integrator who configures the KNX devices and puts the KNX system into operation. The configuration generally involves a few steps; after deciding with the homeowners which home automation functions must be made available for voice control, the group addresses of the KNX system are connected with the syntax provided by the specific home speaker and the final test is carried out.
It can become an important element of the home automation package
The temptation to consider voice control of home automation functions as a mere gadget is understandable. But, on closer observation, there are several factors that speak in support of this new feature. First of all, it must be noted that the control of a smart home has already been greatly simplified thanks to smartphone apps. However, there is a category of users who are not familiar with the technology and in the same household today live people with very different attitudes with respect to the use of the mobile phone. In these cases, voice control offers the possibility of interacting with the smart home in the most natural way possible, without having to go through menus of selection or by interpreting conventional signs or graphic symbols. Then it is necessary to consider the needs of the most fragile users even when faced with the simple daily actions to be carried out in a home; this is the case of people with disabilities and the elderly who are not self-sufficient and whose quality of life could be significantly improved by the use of home automation in combination with voice control. But perhaps the most interesting aspect of voice control is that, in the eyes of end users, everything appears to be natively integrated into the smart home. In reality, this is rarely the case, since the home speaker is interfaced with a multitude of services and products (single or entire systems) that speak very different languages. This "multiplicity", however, remains hidden from end users who have the perception of a single interface between them and the world of smart home. And this is no small advantage: regulating a light becomes as simple as knowing what temperature there is outside and the instrument used is only one: the voice.
Evolving technology
Voice control can therefore represent in the short term a small revolution in the field of man-machine interface in the domestic environment, making more easily understandable systems and technologies that are considered difficult to use by those users who are not particularly inclined to technology. In any case, we must not underestimate some aspects that need to be improved to consolidate the technology:
- The chain of actions required, for example, is quite long: once the sentence is pronounced, in fact, the audio track produced is transferred by the home speaker to the cloud where a natural language processing engine provides for its translation into digital format, resulting in a text file. At this point you must identify the service or device involved, forward the request in a format that is comprehensible and finally give feedback to the end user along the chain in the opposite direction, from the cloud to the home speaker. In this sense, it would certainly be preferable to have the option of local functioning available at all times.
- Finally, in terms of privacy, it is wise to think of an intervention by the regulatory authority to avoid security risks for personal data. It is technically possible to capture and store data from users' private conversations; various aspects must therefore be regulated and users must be assured of the confidentiality of the functions performed and data protection.
- The syntax is still a bit 'rigid and certainly needs more flexibility to be able to perform the action required by the end user. Different ways of expressing the same request will have to be recognised, especially to avoid learning from memory a long list of sentences for controlling the home automation system. However, home speaker manufacturers are constantly working on this matter.