In a hands-busy and eyes-busy activity such as driving, spoken language technology is an important component of the multimodal human-machine interface (HMI) of an in-car infotainment system. Adding speech to the HMI introduces two distinct challenges: accurately acquiring the user’s speech in a noisy car environment, and creating a spoken dialog system that does not require the driver’s full attention.