Improving ultrasonic sensor data

AUSYS developed an automated system using ultrasonic sensors to monitor woodchip levels beneath an industrial crane.

In practice, the irregular surface, crane vibrations and unfavourable sensor angles can produce unstable readings. A single incorrect value may cause an unnecessary stop or an inaccurate level estimate. In this scenario, reliable data processing is essential. Basic processing, such as averaging, can be distorted by outliers. Therefore, it is necessary to separate valid measurements from sensor errors in real time.

OWASmooth removed incorrect readings while preserving the sensor’s actual movement. The processed signal followed real changes in distance and ignored readings inconsistent with the surrounding data.