Today's Advanced Driver Assistance Systems (ADAS) are based on cameras, radar or lidar. As a result, objects must be within the line of sight in order to be detected by the system. This becomes a weak point when it comes to emergency vehicles - they can be heard much earlier than they can be seen and are thus perceived later by ADAS. To overcome this challenge, Infineon Technologies has developed a sensor solution together with Reality AI that gives vehicles the sense of hearing. This solution complements existing sensor systems with MEMS microphones. This allows cars to "see" around corners and warn of moving objects in a blind spot or approaching emergency vehicles that are even further away.
The new sensing solution, shown at Infineon's Virtual Sensor Experience, is based on Xensiv MEMS microphones combined with Aurix microcontrollers (MCU) and Reality AI's Automotive See-With-Sound (SWS) system. Using machine learning-based algorithms, the system is able to detect emergency vehicles, cars and other road users even if they cannot be seen by the driver - or by the sensors used in the vehicles' ADAS. Machine learning also ensures that the country-specific sirens of emergency vehicles are recognised in all parts of the world.
Infineon's automotive-certified Xensiv MEMS microphone IM67D130A has an extended operating temperature range that allows various use cases in harsh automotive environments. The low distortion (THD) and high acoustic overload point (AOP) of 130 dB SPL enable the microphone to capture distortion-free audio signals even in noisy environments.
This allows reliable classification even if the siren sound is masked by strong background or wind noise. This sound-based sensor technology can also be used for other vehicle applications such as road condition monitoring, damage detection or even predictive maintenance.
The software for processing the audio signal runs on Infineon's Aurix TC3x MCU family, which is used in numerous automotive applications. This scalable MCU family offers a portfolio of one to six cores