Audi | Voice Assistent

Strategic foundations and prototyping a vision for the future

Digital voice assistants have moved into countless homes in the last few years and tend to have particularly our entertainment and smart home devices under control.
But what about our second home, our vehicle? While the experience of a voice-controlled assistance is not fundamentally new, it has not really been a satisfying one for the user due to limited commands, poor understanding and slow analyses. However, cloud access, natural language understanding (NLU) and artificial intelligence will change this significantly in the future.

Defining the strategic foundations

In 2018, we supported Audi in the fundamental strategy process for developing its voice assistant by bringing together a multidisciplinary team of business and brand strategists, designers from various disciplines as well as developers. In four collaborative workshops, we were able to define the most important parameters of the voice assistant as well as its personality and its effects on voice, style of speech, sounds, actions and screen content.

skills

Designing the user experience

In 2019 we have turned the vision into reality with a mature prototype: Together with the AI experts from paragon semvox and the software developers from iNTENCE, our goal was to take the in-car POI search to the next level through multimodal interactions based on AI algorithms. Our inspiration was the perfect concierge experience, which quite naturally takes the context and individual preferences into account.
With the vision of a future in which in-car interactions will primarily be voice based in our minds, we developed dialogues, sounds and screen content for various use cases ranging from getting to know basic user preferences and proactive POI suggestions from the assistant to collaborative POI search with a second person. The dialogues scripted together with a screenwriter were recorded by a previously casted speaker – taking into account the personality of the voice assistant and its adaptation to different situations.
In addition, we developed a concept for the essential visualization of the voice assistant in the car using screen animations and ambient light.

flow

Ready for the future

The final result of the collective work was a true-to-detail multimodal prototype, which was piloted at an internal Audi innovation event in October 2019 and whose user experience will hopefully be experienced in series production in the near to medium future.