Generative AI is poised to revolutionize the car cockpit and IVI
(In-Vehicle Infotainment) system experience, making the driving
experience personalized and adaptable to driver preferences.
A free text, conversation-like interaction between driver / passenger
and car has the potential to enhance safety, comfort, and entertainment,
making every journey unique and tailored to the individual.
There are many advantages to running generative AI (GenAI) locally on the IVI
or the infotainment ECU rather than in the cloud. Firstly, GenAI on the edge enables instantaneous, real-time interaction between car and people. Local GenAI capabilities allow for interactive interface with immediate and personalized response to queries and requirements. Secondly, in areas with spotty or no network reception, GenAI on the edge can still function, offering crucial assistance without relying on an internet connection. Finally, privacy concerns are mitigated by keeping user data and driving habits within the vehicle, minimizing the amount of information sent to the cloud.
Running on the edge, GenAI can assist drivers and passengers by enabling voice-based, natural speech control of the vehicle, for operation of the different car systems including infotainment, navigation, and even repair & maintenance, relieving the driver from searching through multiple tedious menus and sub-menus.
The Hailo-10 AI accelerators enable power-efficient, cost-effective generative AI on the edge. Unlike high-end Cockpit SoCs with integrated NPUs, which are very costly and apply only to premium vehicles, Hailo’s Generative AI Accelerators work as co-processors with entry to mid range Cockpit SoCs, making local GenAI affordable for mass market vehicles. The Hailo-10 is a self-contained solution which does not rely on shared resources with the main SoC, ensuring efficient execution of inferencing and Gen AI functions, and guaranteeing freedom of interference for smooth operation of all Cockpit and Infotainment functions.
Prevents costly monthly subscriptions to multiple cloud-based generative AI tools
Local processing reduces latency and enhances user experience
Ensures continuous access to essential tools and services, even if internet connectivity is limited or unavailable
Mitigates both personal and business-related risks associated with transmitting personal or sensitive data to the cloud
