Measurements of the farfield characteristics of an automotive ultrasonic sensor with a vertical temperature gradient

Abstract

With the coming of autonomous vehicles on the horizon, a variety of sensor platforms such as ultrasonic, radar, lidar, and image sensors are currently under development. Among these, ultrasonic sensors play a key role in the autonomous parking system. An ultrasonic sensor emits and receives a sound pulse (usually centered around a few dozen kHz), and computes the distance to a nearby object assuming a constant sound speed. However, the assumption of constant sound speed is often invalid because of the meteorological conditions, and could cause errors in distance estimation. In this talk, we discuss the farfield characteristics of a typical automotive ultrasonic sensor in the presence of a vertical stratification of temperature hence sound speed. The sound speed profile on a hot summer day is derived from temperature measurements with a vertical thermocouple array, and the farfield acoustic pressure is measured using a 2-D array of microphones. The measurements indicate that refraction due to the stratification of sound speed can be significant enough to alter the farfield performance of an ultrasonic sensor, and thus should be taken into account in the design of the autonomous parking system.