Hearing car offers reliable detection of emergency vehicle sirens

26-10-2022 | Infineon | Test & Measurement

Infineon Technologies AG is working with Cerence Inc to create a solution based on automotive-qualified XENSIV MEMS microphones (IM67D130A) from Infineon and Cerence Emergency Vehicle Detection (EVD) software. This will allow vehicles to actively detect approaching emergency vehicles, particularly when they are not in sight.

"As autonomous vehicles quickly become a reality, emergency vehicle detection will be critical to provide drivers with the information they need in emergency situations," said Christophe Couvreur, SVP and GM, Core Products at Cerence. "By partnering with Infineon, a global leader in MEMS microphone technology with a strong presence in automotive, we provide OEMs an integrated hardware- and software-based emergency vehicle detection solution that will enhance road safety around the globe."

"We are committed to collaborating with industry leaders for next-generation innovations. Cerence is a global provider of AI-powered software solutions for vehicles and a great ecosystem partner," said Frank Findeis, SVP and GM, Automotive Sense and Control at Infineon. "We are able to bring Cerence Emergency Vehicle Detection together with our high-performance MEMS microphones to provide automotive manufacturers with a best-in-class solution to meet demanding market needs."

The emergency siren detection system delivers a complementary detection modality for autonomous vehicles and improves their perception capabilities. The system combines an array of XENSIV MEMS microphones strategically placed on the outside of the vehicle with Cerence EVD.

The microphone is qualified according to AEC-Q103-003 and features an extended operating temperature profile from -40C to +105C, allowing various use cases in harsh automotive environments. With a total harmonic distortion (THD) of less than 0.5% at a sound pressure level (SPL) of 94dB and a high acoustic overload point (AOP) of 130dBSPL, the microphone can capture distortion-free audio signals in noisy environments. This permits signals to be reliably classified even when background noise obstructs the siren tone.

Cerence EVD may be integrated into the automotive assistant or on separate controllers, it utilises microphones to accurately and reliably detect the sound of sirens. Also EVD can evaluate the sound source location from emergency sirens of police cars, ambulances and fire trucks. Once a siren is identified, the driver, or in autonomous vehicles, the automated driving assistant, is informed to react accordingly. Human drivers can be notified in a multi-modal way, e.g. volume of the radio or other media is reduced, a visual warning may appear on the head unit's screen, and an acoustic warning is given via the vehicle assistant.


By Seb Springall