Data-driven and objective-based independent testing for DP vessels Summary
We recently had the pleasure of participating in the Data-driven and objective-based independent testing for DP vessels webinar, hosted by Riviera. Mark Craig, President of OneStep Power spoke on comprehensive power testing solutions, along with:
Suman Muddusetti. Projects and Technology - Principal Marine Operations Engineer, Shell International Exploration and Production.
Mat Bateman. Senior Partner/Marine Consultant, Keelson Marine Assurance.
Dag Leo Emblemsvåg. Business Lead Dynamic Position Services, Global Maritime.
This is a summary of the webinar including individual presentations and Q&A from the whole panel at the end. We look forward to the next webinar and hope this summary is helpful. we highly recommend checking out the original webinar and the others from Offshore Energy Week in the Riviera Webinar Library.
Slide from webinar presentation.
Suman Muddusetti started with a presentation on data centric evidence - What is data centricity? He noted that the data should provide evidence for validating the post failure capability of a DP vessel. The results should be easily understandable by anyone who reads them, not just the technician who performed the test. The goal of the results is to “establish a basis of confidence” for everyone that the vessel will operate incident-free. The objective is delivering predictable, incident free DP operations.
Slide from webinar presentation.
Mat Bateman talked about thinking about your data and the Data Four V’s: Veracity, Volume, Variety, and Velocity. He simplified these to Integrity (is it any good, is it reliable?), Amount (where to store the data?), Speed (how often do we need to measure for each situation?), and Difference (there are many different types of data). We should consider: What is the content of a good test program? A successful program meets guidance and trials with findings based on data. There are two hurdles - we need to trust but verify with independent witnesses and as remote trials become more attractive, there is no specific definition for these trials. Second, how do we get the independent witnesses on site virtually? Virtual reality and augmented reality are potential solutions.
Slide from webinar presentation.
Dag Leo presented on what makes a great DP trial. Robust trials that provide an equal or better outcome to current tests are essential. There are some possibilities looking forward, such as technology evolvement that could provide opportunities for remote testing and integration of testing equipment in vessels. Challenges include bandwidth, security, and current hardware abilities. He was careful to point out that new solutions should not be created for the sake of creating a new solution, but to increase assurance. "A new solution would never be better than how it's executed."
Slide from webinar presentation.
Mark presented on: what makes a good test program? We’re all doing this for operational reliability and safety at sea, to do that we need to focus on data-driven solutions. The optimal testing program should back up all results with hard data, without ambiguity, quickly. Instead of using poorly defined test methods “because it was always done that way” look for the systematic solutions that support the 7 Pillars of DP:
Fault ride through
Tests should not change settings on the vessel and should provide results instantly. As Suman mentioned, different people should come to the same conclusions looking at the report. We should collect the data that is needed and use visualizations to present the data as usable information with graphs and annotations that are easy to understand. Remote trials need to be seriously considered, with the biggest barrier being the bandwidth of ship operators in performing these tests. Vessel crew may require some training and adopting a different mindset for testing.
After the presentations, we had a phenomenal Q&A discussion with all participants. Questions have multiple parts and as above, this is our interpretation of the summary of the answers from the panel. Our goal is to present the great information from participants - we believe the information is valuable even if it is not word for word. We recommend checking out the full webinar - it was jam packed with information. Here is a summary of the questions and answers:
The standard and skills of engineers on DP vessels varies hugely. Data-centric may be another step beyond their training.
Typically DP tests vary in complexity - what is the benefit of a data-centric approach when testing a simple alarm. Would it increase complexity?
Suman said there is value in crafting a good quality program based on the FMEA and tied it back to Mark in that the test should be very clear on why we are doing it and what role it plays in defending the redundancy concept. The panel continued in saying that a well-defined test and well-defined procedures reduce the need for specialization of the tester. Don’t overcomplicate the situation and verify the item needing verification. It’s also dependent on the engineers and the specific knowledge they have. Improving the knowledge of the engineers would help as well as eliminating unneeded data that can make analysis overwhelming. However, getting pure data is a bit optimistic in some scenarios.
The second question was three parts:
Data driven is presented as a PDF presentation document. This is the traditional way of approaching results of trials. The digital verification methods developed lately are light-years away from PDFs.
Do the panel believe that the vessel should have all the test equipment on board or should testers bring it on board?
Should DP FMEAs be assured in accordance with the OCIMF framework before testing?
Driving up the quality of FMEA will increase transparency and one cannot devise a good quality trial or program without basing it on a quality FMEA. There is room on newbuild vessels for testing equipment, and at the same time we should rely on people like Mark and the team at OneStep Power to carry out verification. Digital verification is key, the PDFs provided are just a visual representation of the data that is captured. With 4,000 samples per second, there should be data tracing capability, not relying on personnel to verify the data. Newbuilds should absolutely include data logging as it is cheap and readily available technology. When FMEA doesn’t have everything you need, the quality of the test can be affected. Building to test is important, but it’s not currently being done. It’s really needed going forward and will help drive good solutions. Trials verify the FMEAs and if they’re not good, there’s not much use in it.
OCIMF is the Oil Companies International Marine Forum. Their mission “is to lead the global marine industry in the promotion of safe and environmentally responsible transportation of crude oil, oil products, petrochemicals and gas, and to drive the same values in the management of related offshore marine operations. We do this by developing best practices in the design, construction and safe operation of tankers, barges and offshore vessels and their interfaces with terminals and considering human factors in everything we do.” You can check them out at the OCIMF website.
What would the panel accept as data-centric proof of functioning on an alarm e-stop?
Testing guidelines are based on industry standards. Should we be pushing IMO to update current DP requirements to aid legislative requirements inclusive of new technology?
We focus on newer vessels. Older vessels are not equipped with data logging that would support data-centric evidence. Upgrading equipment can be costly - would charters subsidize implementing data logging tech?
Charters are unlikely to subsidize the tech. We should ensure when the requirements are specified so it is priced in, making it easier. We need a uniform understanding of the guidance to make sure it is followed. There is still a way to go to get the correct data from systems, but fully remote trials could get established for newbuilds. We shouldn’t leave vessels behind. Guidance was needed for remote vessels, but we could work within existing documents for remote testing. On the subject of the e-stop: you need to know if your e-stop operates the system and if it operates the way it is supposed to. You don’t need to necessarily connect monitoring to the e-stop, but find out if the system is operating the way it is supposed to when pushed. OneStep currently uses existing guidelines to design our test programs and if we keep adding to them it could create too much complexity. With data logging, costs are dropping significantly, but boats are run to make money and there is still pressure on driving down rates, still leaving the question of who pays for the data logging equipment?
Key conclusions and takeaways
Trial results marked “as expected” is not data centric. We need to know what happened, and that what happened is what was expected. Clear documentation is important. The only objective is that a failure should not change a vessel’s position or heading, but “as expected” isn’t going to cut it. There are too many “as expected” results without adequate data. There is also some skepticism as to whether we are ready for this change.
The industry as a whole is heading in the right direction toward data-driven testing - it gives higher confidence in the ability to keep station after an incident. There are a lot of people out there thinking the same way with improvements over the next few years. Virtual testing is possible, but we need to consider the bandwidth needed and how do we store the ever increasing amounts of data? We need to think about that now so we are prepared.
We are very thankful to be able to participate in the Data-driven and objective-based independent testing for DP vessels webinar and appreciate Riviera for hosting it. Thanks to Suman, Mat, and Dag for providing an insightful discussion and the viewers for their excellent questions. We highly recommend checking out the rest of Offshore Energy Webinar Week.