Victor Seleykov
The Future of Voice on the Go
Ethan Richardson, the Founder of Voiceweave, explains how voice technologies are shaping the future of the automotive industry
Five years from now our vehicular experiences will be quite different from those of today — analysts of Frost & Sullivan predict that the importance of digital voice assistants in automotive branding will increasingly grow. The news from Honda and Lamborghini we heard at CES earlier this year proves this point. So, how did in-car voice assistants evolve from car steering wheel voice recognition systems? Why did they take off? And how are they shaping the future of the automotive industry? We addressed these questions to Ethan Richardson, the Founder of Voiceweave, who authored this piece.
Evolution of in-car voice assistants
In-car voice assistants have been around for a while already: they began as embedded car steering wheel voice recognition systems. Capable of responding to simple commands and phrases, they still contributed to shaping consumers’ habits when it comes to in-car voice capabilities. In 2004, Honda installed voice assistance through IBM ViaVoice.
In 2013 we saw Apple introduce the Siri eyes-free option for CarPlay which kick-started the concept of using embedded voice technology in vehicles. Google soon followed suit in 2018 with the Google Assistant integrated into the Android Auto platform.
The next step for voice assistants was to graduate to third-party devices such as the Roav VIVA and Roav Bolt which first brought Amazon Alexa and Google Assistant, respectively, into cars. Furthermore, some car manufacturers invested in the realization that they could develop their own in-car voice assistant. One of these manufacturers was Honda with its Honda Dream Drive OS first presented as a concept in 2017.
In 2018, more manufacturers started creating their own first-party embedded voice assistants in partnership with specialized speech technology companies. Mercedes-Benz introduced the Mercedes-Benz User Experience (MBUX) which built on the capabilities of the Nuance with their Natural Language Processing (NLP) technology.
From this point, Amazon in 2018 unveiled its own purpose-built device with the Amazon Echo Auto. Although not immediately released we saw consumers’ interest and anticipation grow. Now generally available in the US and India, this device aimed to deliver the Amazon Alexa voice assistant to vehicles without an embedded voice assistant.
In 2019 and 2020, car manufacturers proceeded to invest in embedding Amazon Alexa Auto into cars and some even creating their own voice assistants.
With this rise of in-car voice assistants, drivers increasingly use them. Nearly twice as many U.S. adults used voice assistants in the car (114 million) as through a smart speaker (57.8 million). We are experiencing voice usage increase year on year to a point where now there are more than 77 million monthly active voice users.
Future of voice is inside vehicles
Car usage has steadily grown over the years and so to the advancements in technology. We have seen voice grow as an interface from controlling cabin functionality to now entertainment. The fact that driving requires the use of our hands leaves voice as the natural user interface to perform tasks such as playing music and navigation. In-car tasks tend to be one time commands that require little cognitive load so the voice is the ideal method to use while we can focus on driving.
Voice technology from automatic speech recognition to language understanding has also gotten to a level where it can be reliably be used as a gateway to services on-the-go. Furthermore, the coupling of real-time location capabilities voice now becomes location-aware opening up a new chapter in the development of location-based voice services.
Having the ability to speak with a voice assistant and having it know or location creates a rich, relevant and rewarding experience while we are on the move. This type of application of voice is the future of voice, one where we can use it to perform tasks with our current location in mind. For example, imagine being able to check the stock availability of products in stores around where we are, also being able to order food while on the highway before we get to a service station.
Perils of designing voice assistants for cars
Designing voice assistants for cars requires a new type of approach to voice user interaction design. Voice assistant experiences can not simply be copied from a home setting to an on the go context.
We need to take into consideration the fact that we are on the move and our attention is not one hundred percent on the interaction with the voice assistant. This means that we need to change the way that the voice assistant talks with us, the speed of the voice assistant needs to be slower, the phrases need to be succinct and clear. Voice assistants need to be able to handle the request for it to repeat what it just said as well as the ability to extend the default time allowed for the user to speak.
By simply cloning an in-home voice experience to an in-car environment we risk, at a minimum, causing sub-par experiences with the same voice solution. At worst we risk adding to the number of distractions to the driver.
In-car experiences need to be user-led experiences allowing for the user to dictate the pace of the interaction, all the while, voice solutions need to cater to the change of experience depending on the device being used.
In one particular example, Alexa based voice skills need to become smarter by using the features available such as intent history. Using tools like this can give the ability for the skill to reference past interactions to increase the understanding of actions that the user might want to perform rather than prompting the user for clarification.
In another example, if the user is using a device with location access, such as the Amazon Echo Auto, certain features and or the skill audio response can be tailored depending on the driver’s current speed.
Smart functionalities like these make for a more engaging, enjoyable and safer experience for the driver.
What content providers should consider
As the world becomes more ambient and there are more voice assistant enabled devices coming on to the market we are finding more services and solutions being developed for our audible attention.
With this in mind content and service providers need to consider the context in which audio is consumed in the car. Firstly, the car is a communal space, meaning that this space is more than not a place where there are multiple demographics present. Children could be present in the vehicle, not just adults. From another point, providers need to be aware that voice commerce on the go comes with risks such as privacy. How do drivers perform secure purchases in a shared environment? How do voice experiences create the ability to tailor responses depending on who is using the voice application? How does a voice application tailor the content output as well as suggested content based on who is speaking?
Fortunately, there are new speech recognition technologies currently present that allow voice applications to recognize the individual voice. This could be one way that voice applications could secure access to services, commerce, and content.
Looking at in-car consumer behavior also means looking further at the scenario in which drivers find themselves. Imagine drivers who are commuting from work to home. Presenting services that require immediate action and decisions may not be a welcome experience. Voice applications need to be aware of the context and in this example give the driver an easy way to save away their searches for retrieval at a later date and time.
Furthermore, drivers are on the move while driving so this implicit behavior should be considered when forming responses to the driver. Voice applications should also take this into consideration when presenting visual content to the user.
Most promising CES 2020 announcements
The doubling down by Amazon at CES to embed its voice assistant into more and more cars was a notable and significant effort to bring in-car voice capabilities to life. With announcements by Amazon that the Lamborghini Huracan Evos, along with the Rivian R1S and R1T would be released with the Alexa voice assistant on board was a huge signal that Amazon is serious about the voice on the go and sees the car as the next space for voice assistants.
Furthermore, Honda also signaled its intent to play in this space by releasing its Honda Personal Assistant which leverages the technology of the SoundHound platform. We saw that the in-car space is an important battleground for companies to develop products and platforms and this is one of the most promising things for the voice inside the car.
The announcement that Amazon will make its Echo Auto device generally available in 2020 is another huge step towards making voice assistants pervasive and accessible to all drivers. The very fact that every vehicle can now have a voice assistant opens up the large potential for services to be delivered to drivers on the go.
Potential of voice on the go
The potential for voice on the go is huge especially when it comes to voice commerce. We are no longer bound to the boundaries of our homes making voice on the go an untapped space for retailers to really provide their products for purchase. The key opportunity is hidden in the real-time location.
This opens up a world of context and location-aware services that take advantage of tailoring content and services depending on your current location.
For retailers, this space becomes a viable opportunity to extend their existing mobile presence. They now have the opportunity to take advantage of real-time location capabilities to present their products and services in a contextual way. Imagine being able to book an MOT with a nearby garage with your voice when you realize that there’s a problem with your vehicle. Imagine driving past a theatre and being able to find out about shows at the weekend and then reserving or purchasing tickets using your voice. The list and opportunities are endless.
Companies should be looking at how they can augment this on the go space to easily provide their products and services within a context-aware experience. Getting the experience right in this on-the-go environment is key to being successful so a focus on analyzing and testing experiences should be treated with high importance.
Interested parties should be trialing new on the go concepts to learn and understand about this unique opportunity. The barrier to entry is very low at the moment so it makes sense for companies to spring up to provide aggregated platforms that companies can jump to in order to take advantage of this untapped consumer attention.