The Smart Mooring Integrity Checker (SMIC) system detects mooring line failures even in benign conditions, giving indication of a mooring line failure before a potentially damaging storm condition comes through the field. SMIC achieves this using proprietary machine learning algorithms to understand and describe the underlying mooring system behaviour without the need for complex numerical models of the hull and mooring system. Once this understanding is achieved by SMIC, it can predict the system performance in the prevailing environment and compare this to the measured performance. A difference in predicted performance to the measured performance is an indicator of mooring line failure.

What is SMIC?

SMIC is a hardware, software and support service offering, easily customisable for each floating production unit. SMICs major components are:

  • Rack-mount or desktop computer with network connectivity
  • An interface setup for all input data streams (either in software or hardware)
  • The SMIC software package
  • A monitoring interface, either intranet web based or integrated with existing systems aboard the floating production unit
  • Model training, SMIC system maintenance and around the clock anomaly response by experienced AMOG engineers
  • Regular performance reports on the response of the mooring system

How does it Work?

One of the key characteristics of SMIC is its ability to use already installed monitoring equipment on a floating production unit. For this reason the system is cost-effective and easily retrofittable without interruption to normal operations, especially when compared to other monitoring technologies such as load cells or fairlead angle measurement devices.

SMIC uses key environment and vessel data streams to train several bespoke machine learning algorithms on the performance of the mooring system; typically vessel offsets and frequency domain response. This training generates a model of the system. The system then uses this model to predict the response of a floating production unit to the environmental conditions being seen in field. The prediction of system performance is compared to the measured system performance over a sample time period. If the prediction diverges from the measured performance a hierarchy of response is triggered, allowing investigation of the anomalous response.

SMIC incorporates a resilient combination of bespoke machine learning algorithms, capable of coping with irregular, low quality or missing data feeds for extended periods of time. This resilience improves over the life of the system, with more data improving the model of the underlying system. AMOG's bespoke machine learning algorithms have been developed and trained on actual floating production units using data-streams from systems already aboard. This system training included identification of a real life line failure event.