The allows you to analyze the movement of the smartphone along three axes. For example if the accelerometer records the upward movement of the smartphone and its characteristic turn to the ear the light sensor indicates the preservation of the previous lighting and the proximity sensor notes an increase in the intensity of the reflected light the smartphone receives a signal to turn off the screen. When there is simply no sensor All was well until in a startup offered manufacturers a little savings by removing the proximity sensor itself from this circuit.
To replace it it was proposed to use the set of sensors illumination sensor and accelerometer touch screen microphone speaker and neural network capabilities already available in the smartphone. The new algorithm was named Inner Beauty and first showed itself in Mobile Numbers the Xiaomi Mi Mix smartphone. The role of a proximity sensor with a classic light bulb and photodiode was taken over by an ultrasound emitting speaker and a microphone that captures its reflection. A neural network trained by thousands of examples was used to correctly interpret the reflected signal. The problem is that the neural network simply cannot take into account all the nuances.
In the end it didnt work very well and the smartphones screen could often turn on during a conversation or not turn off at all when it was brought to the ear. With all that the very idea of saving appealed to many manufacturers. And since then many smartphones without a physical proximity sensor have appeared on the market. Among them for example were Redmi S FE Xiaomi Mi Lite OnePlus Pro and OPPO Find X. In some smartphones they decided not to use ultrasound at all limiting themselves to a touchscreen and an accelerometer. Other ways of saving Manufacturers increasingly refer to the proximity sensor according to the residual principle.