Identification of engine order telegraph operational error based on image and speech recognition technologies

The unsafe steering behaviors(USBs) of the crew are the main causes of navigational accidents. Accurate and timely identification of the crew 's USBs contributes to preventing and reducing the occurrence of navigational accidents. This paper proposed a machine learning method for recognizing the USBs while using engine order telegraph in manual steering mode. The standard engine order given by the captain or pilot is recognized by automatic speech recognition technology, and the status of speedometer is recognized using image recognition technology to evaluate whether the behavior and the order matches. Experiment using navigational simulator is carried to verify the accuracy of the method. The results show that the method can contribute to rapid response and recovery from operational errors, which in turn improve navigational safety.