Vessel’s Performance and the fallacies of the Big Data.
Konstantinos Lourandos
Shipping Executive | Innovating in Maritime Sustainability
Vessel’s speed measurement via a Doppler Speed log is implicitly required as per Solas Chapter V Regulation 19.2.3.4 for all ships> 300GT . Measuring equipment should comply with specific performance criteria and as a rule of thumb for digital displays the error in the measured and indicated speed should not exceed 2% of the speed of the ship or 0.2 knots (which is greater). The performance of the Doppler Speed Log is subject to many inaccuracies. The installation location is quite critical and in many cases deviates from the optimum location. In addition there are the following major sources of inaccuracies:
- Zero Offset (“ static” )
- Dynamic Errors due to flow effects (air bubbles etc)
- Water density layers
- Effects of ship motions (rolling, pitching, yawing).
Speed Log is one of the most critical parameters for the assessment of the performance of the vessel.
How does the accuracy of the speed log affects the performance?
In the following practical example we compare the performance of two sister VLCCs (VLCC A and VLCC B) following the same routes. The hull condition of the two vessels is quite the same.
Our Key Performance Indicator is the FOC (tons) vs Vessel’s Speed (knots). From noon data and from regular soundings we have confirmed that both vessels have a similar performance.
We compared both vessels at laden condition (BF<4). The y- axis is the FOC in tons and the x -axis the Speed (knots). Speed data has been auto logged from vessel’s speed log for a time period of six moths.
Surprisingly, there was a big difference between the FO consumptions, of the two sister vessels at the specific speeds that deviated from 10 to 20 tons.
Then, we compared the speed data from the GPS and the Speed log as they were auto logged for a time period of 3 months for VLCC A and VLCC B.
The following graph is for VLCC B
Evidently, there is a steady error more than 10% between the readings of the GPS and the Speed Log.
On the contrary for VLCC A the graph is more logical.
It is obvious that there is an issue with the accuracy of the Speed log of VLCC B, which resulted in erroneous results for the performance of the vessel.
When we compared the FOC of both vessels in laden condition, based on the GPS data we got the following graph, which confirmed that both vessels at the specific period performed in the same way.
I am not going to elaborate further on the performance of the speed log, which is out of the scope of this article. My main objective is to point out on the fallacies related to the performance of the vessel and the great importance of the accuracy of the data.
We are living the advent of the big data era in shipping and we tend to believe that "big" data is good data, regardless of the fact that gathering TB of inaccurate data is useless. It is absolutely vital to focus on the accuracy of our data and the accuracy of our instruments before we proceed with our assessment.
Please bear in mind that in this simplified, practical example we assumed that the FO data as measured by the flow meters was accurate which is also an illusion.
Sales Engineer in Wireless Communication Division at MARAC Electronics S.A.
6 年Very interesting! Allow me to add the human factor when we assume “same route”. Are two different crews operating two sister vessels in the same way?is this factor taken into consideration when data is used for comparing conclusions and performance evaluation ?
Marine technology management | R&D programmes and partnerships | electrical propulsion | green shipping | autonomous shipping | maritime analytics
6 年As a control engineer with Ph.D. grounding control systems theory I am somewhat perplexed by how modern marine control engineering, sensing, and analytics is (willfully?) ignorant of estimation theory and estimators in general. I would very much advise using even simple, qualitative estimators on top of the direct sensor feedbacks from the vessels. As in "no, that ship did NOT just teleport 50 m to the side regardless of what the GPS may be telling you".
Applied mathematician, data scientist, researcher, entrepreneur, consultant
6 年Giulio, not sure I agree the Doppler Speed measurements are useless. What is the yardstick used to declare them totally inaccurate?
On Sabbatical :)
6 年More on Speed over Ground and Speed Through Water.? If you take a containership sailing in the South China Sea, the SoG can be several knots different from the STW.? This is an example of a ship in a strong sea current passage.? So, if you take a passage of a ship in a strong sea current passage, you can have several knots difference from the actual STW.? Now in a case like this, if you are not making the proper hydrodynamic corrections AND looking at both the reported weather as well as Metocean weather, then trying to make an analysis of that voyage is useless - you wont know if you have a poor performing ship or sea current.??