08/10/2026
Cars are becoming increasingly sensor-dense. But do they really need more vision, or do they need better hearing?
Many of the most frequent human-machine interaction moments in a car come down to understanding intent.
Hearing is often the fastest, lowest-compute path to answering those questions.
An audio-first intent layer can provide an always-on “reflex layer” — understanding who is speaking, where they are, and what they mean — while selectively bringing in vision and heavier reasoning when needed.
The result: lower compute burden, faster reaction times, and more robust multi-occupant interaction.
And the car is only where this gets solved first.
As AI moves off the screen and into vehicles, robots, consumer electronics, and smart environments, every physical product will face the same challenge: understanding the people and spaces around it.
Contextual awareness is as essential to Physical AI as vision is today.
Kardome CEO Dani Cherkassky explores the hearing-first approach in his latest article for Mobility Engineering SAE International Media. "Why Cars Need Hearing Before They Need More Vision".
