Recently, AISpeech has announced the mass production implementation of its full-chain voice and language interaction technology in Audi's new generation of intelligent connected vehicles. This collaboration will equip localized models based on Audi's PPE luxury pure electric platform and PPC luxury fuel vehicle platform (such as the Q6L e-tron family, A5L, and A5L Sportback) with an all-scenario, intelligent voice assistant, marking a significant step in their partnership.
The AISpeech-powered system achieves breakthroughs in efficiency and user experience from wake-up response to dialogue Q&A and command execution. It features end-to-end ultra-fast and fluent interaction, supporting instantaneous wake-up, 500ms first-character display, and 1300ms end-to-end response. Combined with predictive semantic streaming, the system begins processing user commands almost immediately after the wake-word is detected, minimizing wait time.
A key capability is streaming multi-intent understanding and execution, allowing the system to process complex commands containing multiple intents in a single utterance. It can handle combinations of high-frequency commands seamlessly, such as adjusting windows, switching driving modes, playing music, and planning multi-stop navigation simultaneously. The "Say what you see" function enables users to directly voice text or targets displayed on the interface for convenient control of settings, multimedia, and applications, enhancing driving safety by minimizing manual interaction.
For reliability in weak or no-network environments like underground garages, the system ensures a consistent experience through offline speech recognition and semantic understanding, aligning core functionality performance with online operation. It also incorporates dynamic semantic VAD (Voice Activity Detection) for natural sentence breaking, adapting to different user speaking speeds and habits for human-like conversation rhythms. Additional features include personalized TTS voice cloning, allowing users to create a custom voice assistant voice from a short recording, and support for Cantonese and English alongside standard and accented Mandarin, catering to diverse user needs in Greater China and overseas markets.
Audi's dual-platform strategy (PPC+PPE) aims to build an intelligent consensus across both fuel and electric vehicles. This collaboration with AISpeech represents a deep synergy in smart mobility, bringing comprehensive upgrades to the voice interaction system in Audi's localized models. The partnership focuses on evolving in-car voice interaction from basic command execution to scenario-based proactive services, integrating intelligent experiences seamlessly into every detail of the user journey to deliver safer, more personalized, and warmer smart mobility solutions globally.