Voice assistants: change and evolution

Posted on: September 27, 2018 by Fabio De Pasquale

The smart assistant market is exploding

According to a research by Canalys, smart assistants represent the fastest growing technological category in history, surpassing virtual reality, augmented reality and even "smart" wearables.

In 2017 alone, over 33 million household digital assistants were sold, mostly Amazon Echo and Google Home Assistant. In 2018, sales are estimated to exceed 56 million.

The role of voice assistants, moreover, is becoming increasingly central within other devices, such as the smartphone. In the US almost one in two people states that they have used a voice assistant, mainly through their smartphone.

The functions of the voice assistants are numerous

In addition to providing information searches similar to those carried out manually by any search engine, they are able to perform more functional activities such as buying on Amazon portals, activating music through Spotify, playing audiobooks and starting videos, movies etc.

In the workplace

There are also professional assistants to be used in the workplace. Coincidentally, but not really if you think about it, the names are very similar to the home systems: Cisco Spark Assistant, Cortana, Eva, Nuance Dragon and IBM Watson Assistant.

These assistants are not secretaries. In some cases, for example, Alexa for Business, this is a jacket and tie version of those for the home. Which compared to the consumer version has already engineered a kind of app store for developers.

The development of artificial intelligence - capable of handling complex requests formulated by users - and the design of interactions with the voice assistant that are increasingly conversational are the key elements for the dissemination of services that exploit voice interfaces.

In the market

Among the markets that could benefit from the combination of voice interaction and artificial intelligence are:

  • e-commerce: among the features that can be activated with Alexa there are already those that allow you to buy from Domino's Pizza, Pizza Hut and Starbucks, even if, according to user reviews, these still need to be improved.
  • Household appliances and utilities: giving voice and ears to the objects of everyday life, such as TV and home appliances. Changing the thermostat settings by issuing a voice command or turning on the house lights is already possible (thanks to Nest), but intelligent objects - able to receive instructions and provide feedback with natural language - are inevitably destined to grow.
  • Car industry: the connected car is already a reality. Ford for Alexa allows you to lock and unlock the doors of the car, check the tire pressure and the status of other components of the vehicle.
  • Banking and security: another fruitful area for voice-based technologies is that of payment systems. Voice assistants can already be used to initiate payment procedures. The long-term goal of implementing voice commands in the banking sector is to use voice as an additional biometric option to strengthen transaction security, alongside face and fingerprint recognition.

The era of voice-first interaction has arrived, but design is not always up to par

Devices based on voice-first and voice-only interactions are destined for an increasingly widespread distribution.

Alongside the expected growth of artificial intelligence - which must increasingly be able to interpret commands formulated in natural language - the design of the interaction must take steps towards better defining the architecture of the conversations between the user and the voice interface and offer an increasingly satisfying experience.

We know very well that this technology is not perfect and that we must ensure that conversations between devices and users are not "broken conversations".

The way forward

This is not easy. Building a natural human-computer conversation is difficult to codify, and for this reason our task, as experts, is to make the human-to-computer interaction equal to the human-to-human interaction.

To do this it is essential not to start programming without having designed the conversations through:

  1. Creating the assistant’s personality, because users will perceive a persona whether you plan for one or not.
  2. Defining the context, identifying potential users to create personas and scenarios.
  3. Making conversation prototypes with bubble speech and translating them in diagrams.
  4. Avoiding broken conversations and finding solutions to an imperfect technology.
  5. Thinking bigger

There is work to be done but we will not stop and the future will be vocal.

Share this blog article

  • Share on Linked In

About Fabio De Pasquale
Digital Business Development and Innovation, Atos
Architect, with a consolidated background in UX and CX, Fabio joined Atos Group in 2015 and currently is part of the Global Digital Business Development & Innovation team for the Manufacturing Industry. He is also part of the Atos Expert Community as Senior Expert in Industry 4.0, Decarbonization and Immersive Experience focused to the Automotive, Discrete Manufacturing and Consumer PackagingGoods and F&B markets.

Follow or contact Fabio