Class Societies aren’t DP Gatekeepers

Whenever dynamic positioning (DP) professionals get together and discuss the poor state of DP vessels, failure mode and effect analysis’s (FMEAs), trials, and operations (Ops) manuals, eventually, the cry goes out, “Why aren’t the Classification Societies enforcing their rules to maintain DP vessel and document quality?”?The answers aren’t really a mystery and are built into the structure of the industry.?Let’s use a slightly weird but fun analogy to look at the problem.

Imagine that criminals got to choose their parole officers (give me a chance, I’m going somewhere with this):

• Parole officer A has a reputation for being strict.?His work is respected and people under his supervision are less likely to reoffend but he is fussy, rigid, and still fallible.?One kid got sent back to the slammer for being out past his curfew despite working with a police officer at a soup kitchen at the time.?He broke the rules and he had to pay.?Police officers and employers generally trust people who have gone through his reformatory process.
• Parole officer B is less strict and respected and his people are more likely to reoffend but he is less fussy and rigid, if more fallible.?He makes mistakes but can often be talked around and the mistakes are just as likely to be in the ex-con’s favor.?Police officers and employers are more wary of people who have gone through his supervision.
• Parole officer C is a really nice, understanding guy because he knows no one is perfect and we all need to help each other out.?His sloppy supervision is least respected and his people most likely to reoffend.?Police officers and employers are distrustful of the ex-cons who have gone through his care and need to take additional steps to protect themselves and the public.

Now imagine that you had choose one of these parole officers for yourself:

• Parole officer A will improve your reputation and employability but his complex rules and uncompromising positions could cost you more time due to your mistake or his.?You might choose him if you really want to go straight and need the structure and supervision to keep you there.
• Parole officer B gives you a decent reputation and employability and is less risky.?You might choose him as a compromise between the other options.?He provides less help but also less risk so he is a good choice if you going straight and don’t want unnecessary hassle (or are staying crooked the smart way).
• Parole officer C doesn’t help reputation or employability but offers almost no risk.?His work officially meets minimum standards until proven otherwise.?You’re a free man and there is lots of work so he might be a good choice if you know what you are doing and need a little more wiggle room to do it.

Employers have a preference but they still need to hire someone to do the work and are usually willing to work with what is available.?The cops are usually chasing bigger fish.?So, the free choice comes down to balancing parole officer risk with market risk and your own capabilities.?In a tight labor market, parole officers B and A are probably preferred, but in a loose one they provide little benefit.

Balancing risk and benefit isn’t the only issue, parole officer A is obviously going to demand a lot more work and time than parole officer C.?The employment and reputation benefits of going with parole officer A or B come with more cost than just approval risk.?Whether the work, time, and risk are worth the result depends on the individual ex-con and the market.

Now imagine that the parole officers get paid according to the number of people that they supervise.?In an environment where there is a lot of work, parole officer A might have few clients and little income.?In an environment where most employers and policemen have poor visibility of the difference between the different programs, the benefits of parole officer A and B shrink and only their costs remain to the ex-con choosing a parole officer.?Finally, imagine an ex-con can freely transfer between parole officers and the new parole officer has to accept agreements made with the old parole officer.?Reputation and employment benefits disappear as the rigor of parole officer A is compromised.?All that is left is the three parole officers fighting to maintain market share, so they can keep food on the table.??

Now, imagine that groups of ex-cons choose the same parole officer together.?Large enough gangs will demand a say in the decision making and their members will be treated differently and forgiven for some faults.?Minimum standards are officially maintained to avoid trouble, but unofficially compromised to keep the lights on.??

When an individual chooses to reoffend, it is obviously their fault, but is the parole system providing the support that it should??When market forces are trumping the parole officers’ protective functions and increasing the rate of re-offense, it is a system problem that needs resolved.?In this case, a structural problem - guardians and the market need to be separate for each to properly do their job.?When you hire the cop who can arrest you, you choose the one who will not give you a speeding ticket.?There are limits to how far the system can be compromised before it becomes obvious and serious pushback occurs, but the process is less obvious in an opaque environment.?Neither the ex-cons nor the parole officers want the judges coming in and cleaning house.?Either the structure or the monitoring of the system needs to improve because self-regulation will not provide adequate results in the described environment.?If self-monitoring was sufficient, the ex-cons could do it themselves.?In the parole oversight environment described, many practically are.?This isn’t good news for employers, cops or the public who face additional risk and risk mitigation burden.

It is probably fair to say that the marine industry has a constant cycle of loosening guidelines and interpretations and then tightening them again after a serious incident.?Looking at the preceding analogy, it is easy to see why.?Governments trust class societies to ensure vessel safety.?Ship owners choose their classification, ships can freely transfer between them as equivalent, and the class societies compete with each other for their business.?This is their bread & butter, and large ship owners have tremendous say in how the class society treats their vessel.?There is only so much trouble that a class society can reasonably expect to make and keep its clients.?At the opposite extreme, owners need to beware of class error, burden, and cost.??

I have worked for owners, operators, class societies, class society consultants, independent consultants, and shipyards, so I know horror stories but can’t tell them because of client confidentiality.?In my experience and the experiences shared by others, there is no role that is uniquely independent and virtuous.?Systems, roles, and attitudes need constantly re-evaluated to avoid the degradation of intended protections.?

We can see similar problems in similarly regulated industries.?The financial industry collapse of 2008 came from government, public, and financial industry maleficence, but the ability of financial companies to choose compliant auditors played a big part in weakening the system by hiding increased vulnerability and poor risk management.?Strangely, this was all known to knowledgeable insiders.?Even outsiders like me knew the danger if not the timing - I turned down a job with a friend that I really liked because of the coming crash in 2005.?A couple years before the crash, one financial CEO explained that they knew the risk but, “So long as the music is playing, you have to dance.”?This isn’t how safe systems work.

It is the constant refrain of many “experts” and agitators that things are terrible and need to change - preferably by granting power and a budget to themselves.?In this case, instead of considering it as pretentious background noise that is demanding power and respect, consider it as a leading indicator of a fault in the overall system.?This is not group think or group polarization of a DP industry subgroup, the problem has real world costs in time, effort, money, and accidents.?Vessel clients, consultants, and industry bodies have good reasons for their DP concerns and have felt forced to create patchwork systems and guidelines to deal with them.?Vessel crews have to make do, and reported & unreported annual incidents illustrate the problems.?DP vessels, FMEAs, trials, and operations manuals that are not fit for purpose and do not meet class standards are regularly approved and this passes the burden of risk mitigation on to the vessel crew, clients, and general public.?Requirement mitigation is not the desired outcome from the classification process.??

To be fair to the class societies, they are under pressure to be more lenient and DP is not the biggest risk that they are managing.?They are setup to capture large and dramatic risks, such as vessel integrity, and DP system risks are down their priority list.?They can’t afford to let anything major and obvious go, but are flexible with “smaller operational issues” like DP.?Expert reviewer, surveyor, and even senior management judgement can be overridden by the need to maintain good customer relationships.?If the structural and engine approval departments have made a lot of trouble then there is little room left for DP complaints, or even sometimes for watertight integrity.?Again, there are many horror stories - stony faces forced to intone “Due to the unique relationship between…” or “In a demonstration of good faith and to assist the project…”?The shipyards and owners aren’t the only ones that need the vessels to get out and work, and to keep working.?Classification society income depends on it.

If classification societies worked for vessel clients rather than owners or shipyards, the pressure would be to mitigate risk rather than requirements, but this change in structure is not a reasonable expectation.?Instead, clients hire consultants and have internal departments to perform the tasks that class did not.?This is a large industry, larger than classification, because each vessel needs to be checked by each client, leading to many duplications of work.?Many owners or operators do not maintain the lessons learned from their previous clients and the cycle begins again.?This is inefficient and less effective than doing it right the first time for all to see.?Clients are struggling to inspect quality back into the vessels and mitigate risks that should have been caught early on.?If the results and costs of these efforts were truly visible, it would lead to reform.

What about the owners??To go back to our analogy, some of the ex-cons are flipping burgers and some of them are performing heart surgery.?The burger flipper just needs bare bone verification while the heart surgeon really needs to be trustworthy.?It’s a poor analogy but the burger flipper and heart surgeon might both be DP2.?To properly mix my metaphors, the burger flipper shouldn’t have the cost and bother to certify himself to heart surgeon standards, but also should not intrude into the heart surgeon market if the burger market slows down (“It’s the same, we both work with meat!”).?Class standards set a minimum and operation above those standards are often not covered.?This is an interesting issue, but at the moment, we are failing to meet the “burger flipper standard” of no single points of failure (or do not steal burgers or organs).?Owners and operators are aware of these problems, and burger flippers and heart surgeons fight it out in industry bodies and guidelines.?One sneaky body even hopes to make money certifying FMEA practitioners in competition with existing legislated engineering oversight (e.g. in the USA/Canada, it is invalid and illegal to perform FMEAs without a P.E./P.Eng.).?Clients are not entirely innocent when they force DP1 tasks to be performed by DP3 vessels.?This forces the burger flippers to pretend to be heart surgeons and confuses the market place.

A simple structural improvement, which should usually be used, is ensuring that the DP documents are prepared for the operator rather than the shipyard.?Generally, shipyards just want to meet minimum requirements and move on, but the operator is then stuck with lowest bidder, lowest common denominator work.?Operators need something better.?Having the DP document provider work for the operator rather than the shipyard better aligns the consultants with the operational needs of the DP vessel and provides a better independent check of the design.?The analysts should get involved early to eliminate problems more easily and inexpensively.?This simple change makes a considerable difference in DP document quality and avoids expensive errors in vessel design and operation.?Good DP analysts should always be aligned with safe and reliable DP vessel operation, but many providers need this reminder.?Providers can be pressured into positive efforts rather than negative results.?Again, there are horror stories to illustrate the difference.?These tendencies aren’t laws, there are conscientious shipyards and uncaring operators, but it is the way to bet.

It comes down to this, simple mechanical rules are almost always better than expert judgement in an environment where the experts are subject to influence.?There is an old saying that where you stand on a subject often depends on where you sit.?This reflects the practical experience that judgement is usually influenced by who you are working for.?This is why truly independent judgement, or judgement aligned with DP operational safety, is so valuable.?Generally speaking, that is not what we currently have.?We either need less human expert opinion, or need to align it to safe and reliable DP operation.?Perhaps the approval process should become more automated so less mistakes are made and external influence is less likely to affect judgement.?Expert systems were already outperforming professionals at limited tasks in the late 80s and early 90s (e.g. heart attack risk and treatment) and well considered check sheets are known to improve human performance.?I know from my time with them that class societies already use check sheets, but perhaps it is time to encode the whole process, so deep learning can remove human error and bias.?Humans will need to provide oversight and it will probably require cooperation between the class societies and designers to create an effective and standardized process that cannot be gamed.?This is a high tech solution that we will eventually go to and modern AI will aid the work as a tool, in the same way that the computer replaced the calculator that replaced the slide rule.

Improved and increased standard structures can reveal problems that might be revealed only by the most tenacious analysts.?It will require safeguards to ensure the information is entered correctly and reflects the vessel.?Properly designed and safeguarded, it will outperform all but the most curious and doubtful analysts and approvers, who dig and dig and dig beyond the surface until they are sure.?Those individuals are rare so a well-structured approach can make considerable overall improvement.?Steps in that direction for DNV were DNV-RP-108, which drew on the FMEA redundancy matrixes developed by me in GLND Americas, DNVGL-ST-0111, which drew on the findings and expertise of Marine Cybernetics, and their adoption of other external improvements, such as MTS guidelines.?These improvements mean nothing if a poorly approved vessel, grandfathered or transferred, has the same class, so DNV has taken steps to differentiate.?Overall DP approval system improvement requires cooperation between the main class societies.?ABS has recently issued new DP rules, but the updates seem to draw on older IMCA documents, and Lloyds’ DP rules are unchanged.?Right now, one class society seems to be more interested in DP than the others.

There is sometimes considerable frustration in the DP community over yet another useless class approved set of DP documents and problematic vessel.?Hopefully, the crime analogy makes it clearer why classification doesn’t pay (-off sometimes in the DP realm).?Class approval of DP requirements is sometimes cursory and leaves a lot to be desired.?They partially avoid blame by assuming that the consultants should have known what they were doing.?If the responsibility really lies there, then the classification society wasn’t needed.?But they took money to review and ensure the suitability of that work, and need to do so better.?Major structural approval system changes are probably not going to occur, but smaller changes can better align DP consultants with operational needs.?This is an improvement but won’t solve the main problem.?Meantime, a lot of money and effort is being wasted to repeatedly fix common and well-known problems.

It is said that Sunlight is the Best Disinfectant.?The best path forward is probably a systematic study of the overall problem to document its existence, extent, and costs, so reasonable solutions can be chosen and applied.?There are a lot of indications of trouble, but no quantitative analysis proving it.?If the problem can be proven and high costs quantified, then actions of reform should follow.?A system reform will probably not be good business for much of the DP consulting industry, which currently provides ad hock patches, but it will be better for the overall DP marine system.?Flag states will probably need to remind class societies of their responsibilities and force system reform.

(Full disclosure – I used to work for DNV, ABS, and for consulting companies now owned by DNV & Lloyds.)

Next Week – DP Consultants