Enamor’s flag system used worldwide for evaluation and optimisation of fuel economy on fleet and vessel levels, has been boosted with machine learning algorithms.
The newest release of the system introduces a number of advanced performance tools making performance monitoring and evaluation much easier. User has to simply define the period of time for which performance shall be evaluated. System automatically employs sophisticated algorithms to provide results in meaningful manner. In order to perform task efficiently some state-of-the-art methods has been introduced:
- Unsupervised learning for data processing,
- Mathematical modelling for base performance identification,
- Data clustering for automatic recognition of operational patterns,
- Sophisticated data cleaning and smoothing in order to emphasize performance trends in scattered data.
A focal point of newest release is the enhancement of performance tools. Vessel-level view has been extended with fully automatic performance trending which takes advantage of collected data and builds reference models without a need for providing additional data such as model test or numerical calculations results. This is a significant improvement for those users who didn’t have information on ship’s performance as designed which is used for building the reference model. Tools already available in ship performance section have been rebuilt and optimised for better user experience and quicker execution.
The fleet section has been upgraded with new tool for auxiliary engines performance analytics. Comparison between vessels in fleet or sisterships is now radically simplified with handy periodic data aggregation. Tools allows for quick identification of performance pitfalls.
User interface has been also upgraded with straightforward definition of time period inputs used by analyses throughout the SeaPerformer interface. User can now much quicker define the required time range by selecting full years/months or particular voyages.
More details regarding the latest changes in SeaPerformer can be found in attached release note or by direct contact with our team.