Smart Questions: A selected transcript of the discussions between delegates on the latest SmartTALK
Our latest SmartTALK proved that the issue of vessel and fleet performance remains as topical as ever. Following presentations form Tom Eversen at Jotun and Daniel Schmode at DNV GL, participants heard a discussion between Maurizio D’Amico from D’Amico Shipping and Craig Eason, Editorial Director at Fathom News, the host of SmartTALK.
While a full recording of the March 1 SmartTALK can be found on Fathom World, the discussion continued fervently once the microphones were turned off. Below is a selected transcript of comments.
On performance data
“I am specifically interested in how these vessels are baselined for your eco-insight system, and what is the minimum sensor suite on-board required to do any effective performance monitoring?”
“All monitoring is always good but how can you turn history to an advisory tool? One example: Historic-speed-power relations can be used to derive base lines for hull performance monitoring. Often Sea trials are done only on certain drafts and speed. Historic data can fill the gaps”.
“I agree with your point, however, how do you account for different stages in the ship-system degradation cycle?”
Why do we still talk about noon data?
“It is good to hear some professional opinions on the subject. A question that was raised during the webinar, that noon data alone is surely not discrete enough to get a good idea of the ship performance. What is the minimum data set required before any effective monitoring can be done?”
“I am analysing a data set at the moment, however it is a mix of manually reported data, 5 minute data on fuel flow, inferred data from met-ocean etc. etc. I can see why there is the use of aggressive filtering.“
“We have made statistical analysis and conclude that for typical question one quality reading per day is sufficient. When increasing the frequency the gain in accuracy (dur to statistical effects) is small.
“It is in my opinion more important to maintain the measurement equipment do avoid systematic bias”.
“In my opinion it all comes down to your need. What problem are you trying to investigate/solve? Is it to increase utilisation (noon reports should be fine) or is it route prediction with AI (5minute data may not be enough)”
“Well I am a special case really, at the moment I am trying to assess the performance of energy saving devices (waste heat recovery etc.) and so having the whole picture becomes essential to understanding the benefits. Otherwise you have a situation where it may appear to be a benefit, only to discover it’s robbing you elsewhere.”
“In your case, you may benefit from doing real time, big data calculations, to derive some preliminary indicators which you could send as more averaged values. With this concept you could gain the advantage of whole picture analysis without being drowned by the data.”
“That’s the idea at the moment, the biggest barrier to the adoption of these technologies at this time is that it is difficult to convincingly prove their benefit, outside of all the marketing and sales blurb from manufacturers”.
“Exactly, data is collected, so have BIG data, but it is always history, and Master needs advises.”
“We combine historic baseline data to check current performance. If there is a performance gap we contact the vessel to identify and eliminate the source.”
“I agree that 5-10 minute data points are more than enough to do the job. noon reports are somewhat irrelevant in the era we are entering where voyage resolution is much greater. Rather than sea trial data and three years of performance across dry dockings for a single ship it is possible to build representative models quite quickly from the performance of a fleet of comparable ships and refine them in operation. Valid trends can then be developed quickly.”
“I absolutely agree, as long as that 5 minute data come with the relevant context to assist in the analysis of course. Particularly with regards to instantaneous specific fuel consumption figures, to allow some direct comparison between main engine performance. Otherwise, comparing a voyage in a force 4 to a force 6 is going to be meaningless.”
“Noon reports should not be used because usually weather changed over 24h period. We recommend snapshot reporting which is a 15min average of speed, power and weather which can be recorded by the crew once a day”
“It’s good to understand how the data is collected as well as how it is used.”
“From a common-sense point of view the 5-10 minute average should be fine for baselining. However how easy is it to account for readings like rudder angle. The autopilot may be over correcting, increasing drag and consumption, but the 5 minute average will not tell that.”
On making the changes you now you can
“It is not a magic tool but at design stage for new building test and retest with designers , for fleet into the water proper training at crew level , at operation department , at tech department level , with help also from IT dep. In my opinion it is the full integration of different approach into the organization that is the tool to success”
“That’s a fair assessment. Improve what you have before looking for magic bullets. Certainly there are huge savings to be had in understanding a ship well and operating it cleanly. Every tonne of fuel saved is profit these days”
“As a side question, since a huge factor in fuel saving is how the crew operate the vessel, is there any interest in incentivising the crew to save fuel when compared to a baseline?
On incentivising crew
“On the subject of incentivising crew. I discussed with a tanker owner that had an internal ranking of ship efficiency. The most efficient ships of the ranking would win a bonus for the crew.”
“Yes something like that sounds effective. Carrot not the stick, so to speak!”
“crew incentive on ship’s performance is tricky: poorly blasted and touched-up hulls with perfectly maintained engine will give the same result as vice versa conditions. A lot needs to be standardised then. By who? Owner or Manager?”
“It can be a bit of a minefield. Especially when self-reported data comes in. That’s where the good baselining come in: But easier said than done.”
“I imagine that most professional crews are in that camp. I suppose there may be more benefit between the charter party and operator, with preferential contract rates. It all boils down to having an excellent understanding of how that vessel will behave in a given condition, which is where the thorough sea trials described come in.”
“Mr D’Amico has a really nice view on the vessel manager of the future. What about the number of vessels they will be able to oversee. Will that increase as data-driven solutions and automation give clearer and quicker insight or contrary will there still be the need for attention to detail and constant monitoring?”
“If an operator needs less time to monitor the vessels, they could be concentrating on other matters.
With few vessel and limited voyage x year is clearly very difficult to get conclusions so a “fleet ” point of view and long track of voyage and long reliable track of data will be the MUST , after an initial trial time 1 or two years than ….A.I. software tools can be trimmed and adopted . But human decision is essential
The feedback loop
“On a slight tangent and going back to a subject you touched during the webinar: using monitoring to prove the real performance of the vessel. Do you think the marine ship building industry will close the design cycle by comparing the actual in-service operational profile with the design one? Will the yards start recording data from the ships they design and build? Will the owners and charterers be happy with this?”
“I cannot give a final answer but , yes in some way due to long relationship with the yard and with the designer via the certification of Class much more can be developed , so to compare series of sister ships built at the same yard all this intelligence will in the future come together …I don’t know under which format and under which legal framework.”