Uncategorized

How Not to Fail: The Titan Submersible Quality Manual (Part 3)

Posted on Jul 25, 2023, Modified : Jul 17, 2023 jshell55

We’re on to the third part of our effort to develop a hypothetical ISO9001 quality manual for the Titan Submersible. For those unaware, this vehicle tragically imploded, costing the lives of the people on board at the time. The company Ocean Gate was trying to charge billionaires to voyage to the Titanic so as to finance their research, the result of which would have been an inexpensive, practical way to explore the ocean floor.

In Part 1 of this we made it through the organizational context, top management, and measurement sections of the ISO standard. In part 2 we tackled the question of resource allocation.

Here is the link:

Titan Submersible Quality Manual Part 1

News Update

In the past few days, even more recordings and transcripts have surfaced to the effect that the passengers and crew of this vessel knew they were doomed. There is a link below. I, frankly didn’t have the stomach to listen to it. If it’s on TikTok it might be fake anyway.

Summary of Parts 1 and 2

Well, here is what we found out. Based on the website information, we figure that the founders of the company were not especially customer focused. They were heavily mission focused, and had an underlying want to be somewhere between Jacques Cousteau and Captain Kirk.

Furthermore, we’re seeing a fairly obvious issue with checks and balances. The boss, is the main fund raiser, plus wants to be the pilot. The number two person involved with the project also wants to be Captain Kirk.

What is missing is a process by which someone can say “no.” Example: There is agreement that the project meet some engineering threshold before human life is endangered. They run a lot of tests, and there is a lot of question. They run full scale missions many of which have known problems.

No one has the authority to pull the plug other than the boss.

We can also theorize, that unless there is a deep pocket investor or two, the project is going to be chronically short on money. As a result, there will be resource allocation issues. Furthermore, the resource by which “outside experts” are brought in to weigh in on the design may have been ignored.

I’d love to go through their documentation system, and record keeping, which an ISO9001 certified system will have. However, I have suspect that this system will also be “light.”

Operational Control (Clause 8.1)

Here is the story. You know that before a commercial airline flight takes off, the Captain does a walk around and then has a thick book of pre-flight checks.

The reason for this is that he is the sole operating authority as to whether the plane is ready to go or not.

The plane itself has a history. There is a design, there is testing to make sure it all works, and there are countless hours of flight tests in which the engineers come up with contingencies in case there is a problem.

On flight day, everybody that has access to the equipment has some sort of training, and anything that touches the plane that looks like a tool has to be recorded, including the qualifications of anybody who was holding it.

During the flight there is continuous monitoring, including GPS, sensor data and recording data of what was going on in the plane.

That way if there were any surprises, the information could be re-incorporated into the procedure manual and pilot training and that kind of thing.

If necessary, the third party organization that runs all of this, the FAA, has the power to ground aircraft until the problems get ironed out.

There are audits all the way through the system to make sure everyone is following procedures, and the resultant records are being kept.

Operational Control

So did any of this exist as it applies to this submersible? If I were assigned this project, I would build such a system, and enforce it to make sure the participants followed it.

The whole thing would be captured in a document that we would call the “quality plan. ” We would refer to the individual processes of design, repair, maintenance, operations and post-flight activities.

Did our adventure-seeking pilot and his partner understand the need for this? Well, let’s just say there is some question, which will be answered by a lot of lawyers and a jury at some point.

Sales and Contract Review (Clause 8.2)

Sales and Contract Review procedures depend on one question, which is, “who decides what the product is.”

In this case, it’s “the organization” that is defining the service. They’re giving you a ride in an experimental submarine, made out of carbon fiber. They’re supposed to take you to the Titanic, and drive around for awhile, and then resurface safely.

So that’s the service. That is what is supposed to be in the scope, which we looked at awhile back.

In this case, the customer communications are under scrutiny. This includes promotional information that they have plastered all over the internet. What they’re doing is representing the product characteristics to some “market.”

From what one of the articles says, the company has represented that it has worked with Boeing to which Boeing has now denied any connection. Side point, the team has evidently purchased some expired carbon fiber from Boeing at an “Extreme Discount”. We will talk about that a little later too.

What I am trying to say that “the organization” needs to carefully control marketing communications in order to keep from getting into trouble. Most professionally run organizations that care about this have an approval process by which information is verified and validated before it gets to the marketplace.

And, this can and would have been validated by a third-party auditor whose job it is to check the company’s website and confirm that the statements made represent reality.

Product Definition and Contract Review (8.2.2 and 8.2.3)

What the ISO standard is interested in is the point at which “the organization” makes a commitment to provide the service.

This “commitment” can take many forms, and usually describes “the product” in great detail.

I’ve posted a “cruise ship passenger contract” below, just as an example. But, that example is not far from reality for this project.

The contract specifies various terms and conditions, acceptable behavior, not bringing fireworks on board and that kind of thing on the part of the passenger.

It also commits the “organization” to at least attempt to give you a pleasant cruise experience. So the process of what gets into this contract is the “contract review” on the part of the organization. Whatever level of the organization responsible for this has to “approve” it, and if there are modifications, that also has to be controlled.

Your third party auditor will make inquiries as to what the process is for this.

The Titan Submersible Waiver of Liability

In this case you’re probably looking at the 4-page “waiver” on this, namely the “waiver of liability.” which I have linked below.

The “Organization” is communicating to the potential customers the risks, including various ways to die, and inconveniences associated with the service.

They also openly state that only 13 out of the 90 dives that this piece of equipment did ever reached the depth of 3800 meters, which is an 86% failure rate.

So “the organization” tried to communicate the danger to the potential passengers. Some fraction of these people chickened out and bailed on the project once they did their due diligence.

.Here is a quote:

The submersible was not registered with international agencies, nor was it classified by a maritime industry group that sets basic engineering standards. OceanGate has said this was because it believed Titan’s design was so innovative it would take years for inspectors to understand it.
The Guardian

The Guardian

Design and Fabrication (Clause 8.3)

The ISO standard says that the organization shall have a Design and Development Process.

The intent of this process is to set up “design criteria”, which is to say what the vessel needs to be able to do, and at the end of the day, have proof that it “works.”

It also makes “the organization” develop tests and other methods to prove all of this prior to release to the marketplace.

In this case, in addition to the design process, there is a fabrication process by which the actual vessel is built.

Design Criteria:

Criteria	Method of Measuring	Success or Fail
Suitable for passengers	Known methods of life support etc. for deep sea vessels	Success (apparent) except uncomfortable and no restroom.
Steerable	Some set of navigation tests etc. that demonstrate vehicle control	Very questionable. GPS failure and also blackouts of communications
Able to dive to 3800 meters without implosion	?	Fail.
Re-useable	?	TBD
Capable of being serviced by conventional surface vessels	Test launches and other activities	Evidently OK
Capable of emergency evacuation	?	Apparently there was no way to blow the main hatch from the inside, so as this vessel
Cheap and Easy to Launch	Need to transport on large vessel	Fail

The Space Program might be an analogy in this case.

The high level project, set up by JFK, was to design a space vehicle capable of launching people to the moon, and returning them safely.

The people that designed this process, including Werner Von Braun, had the high level plan, and then did various design stages in order to arrive at a vehicle which could do the job. This was the famous Mercury/Gemini/Apollo and the project did not advance unless the expectations of the previous stages were met.

This included engineering, life support, mission planning and that kind of thing. There was a high level design plan, testing, and a disaster with fatalities. Eventually, after several incremental attempts, several moon launches were (apparently) made, the last of which was done in 1972.

The Titan Design Plan

There is an interesting article linked below from The Smithsonian which describes the “development process” for this vehicle.

There were several prototype models, the last of which was the Titan, which was developed in the late 2010’s.

The main idea was started by the boss, who is the classic “driver” and wanted originally to be an Astronaut. I will allow the reader to go into the nuances of this.

Here is one thing that I have observed a couple of times. When the designer of something, and the “approver” of that same thing is the same person, it tends to get approved. There is a conflict of interest between reality and the ego of a very bright engineer who spent a lot of time working on something.

So the intent of the ISO standard design clauses is to limit that, or at least encourage the organization to put some control over it.

The standard encourages “the organization” to develop these criteria, come up with objective tests, and then keep records to the effect that the project met the criteria. Is this flexible enough to allow “the organization” to go back and change the criteria to meet reality? Sure. But that decision also has to be documented. At the end of the day there is a set of expectations, and a set of actual results, and the third party auditor’s job is to confirm that the process is met.

Ideally there is a knowledgeable “design approval committee” or something similar to push back on the CEO who is also an engineer, and insists that “diving to 3600 meters is okay” even though the original spec said “3800 meters.”

Purchasing (Clause 8.4)

Here is the idea on purchasing controls:

There are two sets of materials for this, or I should say materials and services. There are materials and services related to the actual construction of the vessel. Also, there are materials and services related to operations. This would be renting the ship to drop them into the ocean, the oxygen and other gasses and energy related to operating the vessel.

There may be other activities that are outsourced.

“The organization” presumably wants to use “controlled” materials for the construction of the device. In theory, the components of this vessel should be traceable back to their original manufacturer. There should be some evidence to the effect that the materials passed their qualifications testing.

Carbon Fiber Questions

Instead, one of the articles below described use of carbon fiber material “received at a discount” from Boeing.

The drawback of this, of course, is the question of whether the material was “good.” Ironically use of “dumpster diving” to get cheap materials is contraindicated in the Aerospace industry for this exact reason.

So there has to be a process to select reputable suppliers who provide prime grade materials. If traceability is a requirement, all the better..

As to the actual purchase order: There has to be sufficient information regarding the product to assure that the specifications are being met, including shelf life if any.

Ultimately, there needs to be a set of records that confirms all of this, including records from the supplier that says the product met the spec.

So if we were to set up a purchasing procedure, and train a purchasing agent, we would have an approval process for suppliers and items that reflected this. We would also have a “counterfeit parts” and “suspect materials” protocol. In this day and age, counterfeiting of components is a thing, even in Aerospace.

Service Provision (Clause 8.5)

I think we indicated above that there is an airline analog for this.

There are pre-flight checks and requirements for the vehicle. Also, there are protocols for determining and reporting repair situations. The Captain does his walk around in the snow in Akron, with a flashlight, before he takes off.

The responsible party, namely the pilot, is in charge of enforcing this. Are there imperfections in this as a system? Yes, probably so, as we know. In NASA there is a committee of launch managers and inspectors that determines everything.

Everything has a procedure. Passenger loading, shutting the doors, launch, guidance, and everything else.

During the operation there is a manual “what to do if you have problems.”

There are continuous recording devices and other monitoring methods. There is ongoing communication with the tower in the destination.

SOP Example from the Navy

For sea-going vessels, this is just the same way. I linked below a 109 page SOP for something the Navy calls “Large Sail Training Vessels. ” These are those sailboats that the Naval Academy launches. This was very interesting in that it spent a lot of time on command structure. This includes who the responsible party is.

In one of these things, the main job of steering the boat rests with the XO who is the Executive Officer. The “skipper” or “coach” has overall management responsibility. The Executive Officer makes the decisions on the fly, and keeps track of operations. At the end of the exercise, there is a “debrief. ” This starts starting with listing all of the errors that were made by the captain and XO.

I’m ready to say that in the Ocean Gate system this was a problem with the organization. The wrong person was driving it. The boss is supposed to have a natural CEO role, make decisions, and advise the XO in case of issues.

What is going on here is that the boss decides everything. He was driving the boat because that is who is was.

Here is a quote:

The Skipper/Coach who is responsible for STC safety must exercise this command prerogative when required to avoid danger(e.g., collision, grounding, personnel safety, etc.). However, this policy and the core of the program itself requires restraint, patience, coaching and the creation of an atmosphere in which the VOST Midshipman Skipper or OSTS Midshipman Watch Captain has authority over the boat and crew, as long as he/she is exercising the requisite judgment, leadership and seamanship to address the situation satisfactorily.

If I were running an undersea exploration vessel, I would at the very least plagiarize most of this, and adapt it to the circumstances. This would include pre-and post launch checklists and inspections, which could be used as a record of conformity for initial inspections of the vessel.

titan submersible quality manual would draw heavily from the Navy training ship

Product Release (Clause 8.6)

In this case, since this is a service process, this might be some kind of “post-trip” documentation. I, the undersigned do accept that Ocean Gate did provide the service satisfactorily.” and a signoff by the guest. That way, there is no question as to the nature of the customer experience.

There is an interesting article below from Aviation Safety that gives some suggestions on this. Once debriefed, the organization can learn from the experience. They can have an ongoing record of the state of mind and body of the passengers at the time of the ride.

Of course there won’t be one in this case.

Control of Non-Conforming Output

The ISO standard states that if there is a non-conforming output (i.e. “screwup”) there is an obligation on your part to evaluate it, and if necessary fix it.

And, once you do fix it, you need to update the “risk assessment”. We did this all the way back in the beginning. They need to conduct training to make sure it sticks.

From what one of these articles says, the vessel was being controlled by a video game controller. At one or more points in the past there were failures in the control system. At one or more points in the past, the launch ship lost the GPS signal of the vessel for a length of time.

Did any of these result in a change in the operating system? We would have to look at the records, which are required in the system, to find out. It is possible that the same person in charge of identifying the problem is the one who has to do the work of fixing. You’d think there was a natural tendency to overcompensate, but not necessarily.

What if there are no records? Well, that’s problematic. Also the ISO standard requires whoever made the determination that everything was OK also sign off on the fix. The determination of “action” if any needs to be attributed to a human or group of humans.

So this is another apparent potential weak spot. Failure to address a known problem.

That fits our definition of “entrenched mediocrity.