People are latching onto the logitech controller, but I think it's low on my concerns list. Controllers are relatively robust, and they stated they had backups, and could surface without any control input by manually dumping the ballast.

Much more concerning are the potential flaws in the carbon fiber hull, as disclosed during the David Lochridge lawsuit (mentioned in another thread), the rumoured lack of surface radios, the limited communication with the surface, the lack of any exits, and even the use of carbon fiber (which is strong in tension, good for high pressure tanks, but weak in compression -- not what I'd want for a submarine under massive compression)

Just to nit pick, what good is an exit going to do at 12,000 feet?

The exit would've been for if they surfaced (that plus a life raft presumably)

When you have a fire and need to surface, it's going to do a lot of good. One of the Apollo crews burned alive because the ground crew couldn't unscrew the hatch fast enough during a pre-flight checkout. One of those astronauts, Gus Grisham, nearly died a previous time when the explosive bolts on another capsule fired early which caused the capsule to take on water.

The problems here are just several fold: maybe the door won't help you at depth, but there's a pretty wide envelope of conditions in which a fast escape hatch is the difference between surviving a fire and burning alive.

The CEO of OceanGate said he "learned the lessons of aviation", but this is obviously BS if he's locking himself in a capsule with flammable equipment, piping oxygen in, and just hoping it doesn't burn. That's a recipe for disaster we know how to avoid.

Currently, even if they could surface, they couldn't get out of the vessel because it's locked from the outside

Actually the wireless controller points to a bigger problem like, I think the underlying reason for using a wireless controller may have been because the computers were inaccessible when the craft was under way. It's not so much the device itself, but if there was a need to use a wireless device that would point to a design flaw.

I mean, there’s more egregious flaws, but I believe this thread was about the controllers?

If we’re talking worst mistake I think the lack of any form of communications equipment is it.

Hell, have a 4km long cable. Even if you can’t communicate because the sub is on fire or imploded you can drag them back up.

Of course it’s not nice if your sub implodes, but that’s kind of a requirement for using it in the first place.

You can have safety, or you can have a submarine tourism startup. You can't have a submarine tourism startup staffed by cautious risk averse people.

How many people should we sacrifice to get a submarine tourism startup?

I believe that question has been answered.

We’ve had one, yes. What about the second submarine tourism startup?

n-1

The use of a wireless input is a bit suspect, but the headline seems rather sensational to me. What’s wrong with using an inexpensive logitech component? The critical systems in my car are filled with cheap off-the-shelf components.

> cheap off-the-shelf components

Your car is filled with cheap off-the-shelf *car* components. There's far more scrutiny given to car parts, even though they're cheap stuff.

But to get back to the controller, while it's not necessarily the worst problem, it is emblematic. That's why people reduce it to that. It signifies immediately what kind of company they were. Emphasis on were.

I think the claim that you’re getting better safety outcomes from a $2 ISO26262 IC over a $30 ISO9001 piece of consumer hardware is a bit spurious in this case. Especially since the operator could just replace the consumer hardware input device with a spare in a few seconds if they ever needed to. I don’t think this is emblematic of the problems with this company at all, because this specific complaint is simply sensationalised and overblown.

It's pretty easy, just hire one subject matter expert (SME) on submarines and listen to them.

So many problems with this scheme, though. No rigorous test (depth, survival, atmosphere), lack of SMEs, carbon fiber hull, failed to learn lessons from aviation (don't lock people in a room that can burn), lack of ventilation, failure to provide emergency breathing, no voice comms, and finally the comms loss during testing that was normalized as operating procedure.

All you'd have to do is not make one of those mistakes, and you'd be doing a better job. These are all issues that are solved on US Navy subs, where we train hundreds of people every year to work on and retire just as many. You don't need to know all these systems, though, just find a retired sub chief and have them walk you through the safety systems.

One thing I've picked up working in software, is that you can enter new domains with relative ease if you just align yourself with experts, and at least listen to their advice. This isn't something that happened on the sub, and I'd be surprised if it ultimately didn't contribute to the total vehicle loss.

That "retired sub chef" mais not be inspirational enough.

https://twitter.com/stillgray/status/1671372833979785217