19.11.2021 Views

HANSA 05-2017

Special Focus: NorShipping 2017 | HullPic Review | COMPIT Preview | Leonhardt & Blumberg | Sewol  salvage | Tugs | German Banks | Tanker | Maritime Politics | Offshore Tender

Special Focus: NorShipping 2017 | HullPic Review | COMPIT Preview | Leonhardt & Blumberg | Sewol  salvage | Tugs | German Banks | Tanker | Maritime Politics | Offshore Tender

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Schiffstechnik | Ship Technology<br />

At recent HullPIC conference users and developers of hull performance<br />

monitoring software discussed how to include the future technology to do<br />

a more effective job. By Volker Bertram and Thomas Wägener<br />

Volker Bertram (DNV GL),<br />

co-organizer of the event, rates<br />

the conference as a success. The<br />

number of participants rose by 35%<br />

compared to the premiere event last<br />

year in Marorka, Italy, he says. At<br />

this year’s HullPIC in Ulrichshusen,<br />

Germany, about 110 international<br />

maritime experts from all over the<br />

world discussed how a ship’s hull<br />

can be further improved. »The high<br />

number of participants shows that<br />

the topics that we talked about are<br />

interesting for the maritime industry,«<br />

Bertram reasons.<br />

Many of the participants ccoming<br />

from about 20 different countries<br />

were from Scandinavia, but also<br />

from the Netherlands, Asia, Singapore<br />

and the USA. For nearly all of<br />

them the network character of the<br />

event was at least as important as the<br />

topics that were discussed. For Søren<br />

Hattel (Force Technology) it was interesting<br />

to see »that other companies<br />

have the same issues as us.« For<br />

Bjarte Lund (Kyma) it was also good<br />

to see »what the competitors are doing«.<br />

Furthermore he points out the<br />

great diversity of topics that were<br />

discussed at HullPIC.<br />

The conference ended with a forum<br />

discussion that had – somewhat<br />

cheekily – been given the title<br />

»Operators and Developers – Galaxies<br />

apart?« To make it short, the<br />

answer to this question appears to<br />

be »no«. On most points, the operator<br />

representatives (Rory Kennedy<br />

from cruise ship operator Royal Caribbean<br />

and Mike Servos from Tsakos<br />

Columbia Shipmanagement, operating<br />

oil tankers, bulkers and containerships)<br />

agreed with the developer<br />

representatives (Manolis Levantis<br />

of Jotun using a simulation-based<br />

approach and Matti Antola from<br />

Eniram using a machine-learning<br />

approach). Moderator Michael vom<br />

Baur (<strong>HANSA</strong>) tried to sound out<br />

differences using pointed remarks,<br />

but the panelists just wouldn’t bite<br />

– they remained true to the atmosphere<br />

of HullPIC where sober assessment<br />

of factual diffculties<br />

was the rule leaving little space for<br />

catch phrasing. Still, a clearer picture<br />

evolved and also the forum was<br />

unanimously seen as a success.<br />

The forum discussed first the key<br />

question that was on the mind of<br />

probably most of the HullPIC participants:<br />

Where are we with ISO<br />

19030? More specifically, vom Baur<br />

prodded the operators: In view of all<br />

the uncertainties presented at the<br />

conference (and there were many),<br />

how useful is performance monitoring?<br />

Mike Servos brought it to a<br />

point: »Five years ago we had nothing,<br />

now we have at least something.«<br />

Rory Kennedy agreed, taking a positive<br />

angle on the state of the standard<br />

and performance monitoring in<br />

the industry. RCCL has started with<br />

performance monitoring some years<br />

ago and made already some impressive<br />

progress with double-digit savings.<br />

It is a process that will continue,<br />

both on the development and on the<br />

implementation. So – surprise – the<br />

users think better than the developers<br />

and the maritime world in general<br />

would have thought.<br />

Is machine learning the way forward<br />

or is it simulation? Perhaps it<br />

was gentlemen behavior, perhaps<br />

simple realization of the intricacies of<br />

performance monitoring: The developers<br />

of the more simulation-based<br />

approach recalled the importance of<br />

calibrating models against in-service<br />

experience and the developers<br />

of the more machine-learning based<br />

approach mentioned the usefulness<br />

of virtual sensors and good hydrodynamic<br />

models. As in HullPIC 2016,<br />

the maritime industry was once<br />

more reminded that all systems on<br />

the market are »grey«, using hydrodynamic<br />

modelling with some system<br />

identification.<br />

The discussion then turned towards<br />

a theme that had woven like a red<br />

thread through the conference: sensors,<br />

human input and how we can<br />

reduce errors in the input data for<br />

performance monitoring. »With all<br />

the discussions about speed logs,<br />

torque meters, ambient conditions,<br />

etc. – should we focus on getting<br />

black boxes or should we focus on<br />

the ›human factor‹?« And touching<br />

on autonomous technology, should<br />

machines do the job (of collecting<br />

and monitoring) or humans?<br />

On this point, all panelists and<br />

quite a few members of the audience<br />

commented, but the positions were<br />

surprisingly close again. The consensus<br />

was that it was not a question of<br />

»either – or«, but »both« played a vital<br />

role. All operators agreed that having<br />

transparent and timely feedback<br />

to the crew improved motivation and<br />

data quality. But data frequency and<br />

good algorithms are important, too.<br />

Daniel Schmode (DNV GL) expanded<br />

on the theme of his paper on how<br />

to reduce errors in performance monitoring:<br />

The crew is key to get bias<br />

(= systematic errors) down; data frequency<br />

is key to get noise (= random<br />

errors and scatter) down. In the end,<br />

you need both for good performance<br />

monitoring.<br />

For this point, the discussion went<br />

naturally to »Big Data«. It is coming,<br />

but as such nobody seemed to be overly<br />

impressed. »A lot of data still does<br />

not equal a lot of insight« received a<br />

lot of agreeing nods both from the operators<br />

and the developers. But using<br />

smart filters to automatically identify<br />

wrong data was seen as a likely way<br />

performance monitoring will evolve.<br />

Information fusion combining various<br />

sensors, on-line services (e.g. for<br />

weather conditions or AIS speed and<br />

course data) and human data reporting<br />

is on horizon.<br />

In summary, the forum discussion<br />

took a rather positive view of the ISO<br />

19030 and the state of performance<br />

monitoring. Both the standard and<br />

the implementation in the industry<br />

are deemed to be work in progress.<br />

The way forward is cooperative sharing<br />

of experience.<br />

M<br />

Photo: Jotun<br />

<strong>HANSA</strong> International Maritime Journal – 154. Jahrgang – <strong>2017</strong> – Nr. 5 77

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!