Ultrasonic and VCSEL sensor fusion for distance measurement in parking assistance

Parking assistance is a desired and common feature of vehicles manufactured in the present years. For a vehicle to offer valuable information for such a system, in most of the cases, ultrasonic sensors are used to measure distance. Being dependent on a single type of sensor can result in an unreliable measurement, especially if that sensor is in an environment that affects its measurement capabilities. This paper proposes a combination of an ultrasonic sensor and a light-based measurement of distance sensor. The focus is on overcoming the technological weak points of each individual solution and obtaining more accurate measurements through sensor fusion. The fusion is based on the usage of an individual Kalman filter per sensor and an external fusion block that uses the Kalman filter outputs. Apart from being presented in graphical form, the results are analyzed using mean absolute percentage error as a relevant metric.