Conscious Cars – Crafted Conversations

This session is a deep-dive into the special challenges which confront programmers and conversational designers when architecting in-vehicle speech interfaces. We will give an overview of the different cognitive needs of drivers as compared to users of other voice-driven applications. After that, we will discuss the specific use cases and associated syntactical and semantical approaches for delivering the right experience. Lastly, we’ll give a glimpse into the future, where conversational AI will play a dominating role over conventional speech architectures. In particular, we’ll touch on the challenges of transitioning from rule-based intent matrices towards LLM-driven intent recognition.

Most drivers (and almost all consumers) use speech interfaces every day.
This session is quite innovative, because it offers a detailed inspection “under the hood”, which gives insights into the complicated processes and expertise necessary to provide effective in-vehicle speech applications!
Insights will be given into:
– The specific role of the in-car assistant
– How vehicle outputs must adapt to widely differing contexts
– Multi-intent management and AI speech architecture
– Types of vehicle prompts and their special challenges