ELNAV - Helm Order Monitor System for Maritime Navigation
The Helm Order Monitor is a sophisticated system designed to enhance maritime navigation by integrating automatic speech recognition (ASR) with ship sensor data to prevent navigational errors. Engineered with simplicity, the system provides audio-visual alerts to bridge teams if helm instructions are omitted or misinterpreted. It utilizes an advanced Speech Emotion Recognition (SER) feature developed in collaboration with audEERING® to detect elevated stress levels that may affect decision-making. The technology includes type-approved components for marine environments, supporting 24/7 operation with a durable multi-touch interface and low power consumption. The Octopus microphone array, a development platform by Fraunhofer IDMT, enables effective far-field voice capture essential for navigating challenging auditory conditions on a ship's bridge. With IEC 60945 and IACS E10 certifications, this system optimizes the recognition of maritime phrases and reduces communication barriers, exemplifying a substantial positive step in maritime safety by addressing stress and language issues prevalent in the industry.
Problem
In an investigation report of an accident that involved a helmsman applying the wrong helm, the Australian Transport Safety Bureau stated: THIS TYPE OF ERROR IS NOT UNCOMMON.
Exxon Valdez
In 1989, the Exxon Valdez oil spill was one of the biggest environmental disasters in history. It resulted from insufficient helm orders' monitoring.
Costa Concordia
Did you know that disaster of Costa Concordia happened due to the helmsman applying the helm on the wrong side shortly before the impact?
Solution
The Helm Order Monitor integrates automatic speech recognition and data from the ship's sensors to alert the bridge team about potential errors. Additionally, it utilizes speech emotion recognition to identify and caution against elevated stress levels on the navigational bridge.
Computer Components
Type approved for Marine applications, PCAP multi-touch, designed for 24/7 usage, wide temperature range, low power consumption, fanless operation, ambient light sensor, integrated buzzer, NMEA0183 input, IEC 60945 and IACS E10 certification.
Microphone
The "Octopus" acoustic sensor front end is a research and development platform, developed with the Fraunhofer IDMT's Oldenburg Branch for Hearing, Speech and Audio Technology, with a far-field voice capture custom-made integrated SCH microphone array, ceiling-mounted and optimized for the navigational bridge.
Automatic Speech Recognition
For a Helm Order Monitor technology demonstrator, Fraunhofer IDMT in Oldenburg, Germany, is providing research and development for the optimal acoustic setup, and a special in-house developed speech recognition system optimized for the recognition of maritime phrases.
Speech Emotion Recognition
We are collaborating with audEERING® in Germany to create a specialized Speech Emotion Recognition system. This system will alert the navigational bridge team about elevated stress levels, a significant factor contributing to navigational incidents.
Alerts
Visual and audio warnings if the order is being omitted, misinterpreted, or wrongly executed
Bridging the gap
Implementation of automated alerts otherwise not available in manual mode steering
Analytics
Transcription of recorded data as an analytical tool/performance benchmarking
Speech emotion recognition
Will warn about unsafe stress level, significant contributor to navigational incidents
Situational awareness
Assistance in monitoring the execution of the orders when the concentration drops
Language barrier
Clear English pronunciation reduces the burden on the listener