Hacker News Reader

Comment by cpleppert | Hacker News Reader

The Xbox controllers are used to control the periscope which is not a safety critical device. Regardless, the navy uses wired controllers and did extensive testing and verification. This outfit didn't do anything like that; in one video with a journalist the bluetooth controller was a 'feature' because they could pass it around the sub.

View on HN

The testing and verification is key. It might even be they were trying to lean on some work done by the navy on the Xbox 360 controller and that got switched along the line for the Logitech, losing one of the main reasons for using the original choice.

In any case, I would hope they brought a spare (or had an alternative method to drive, even if cumbersome), as easy spares is one of the selling points of COTS parts (and long as you verify it's the real part and isn't a revision that looks the same invalidating your testing).

I wonder if using an XInput controller has a perk in that it’s relatively straightforward to find a second source if needed. Or, if one manufacturer isn’t working for them, they have a specification for controlling the periscope.

I'm just imagining it running out of batteries. Then the user non-chalantly asks the pilot/guide for the spares. But they get a blank look. They repeat themselves. They must not have heard. They get a grimace this time and they suddenly realize what a precarious situation they were in all along.

They had spare controllers. Part of the idea of using off the shelf components like this, is the ease of replacement. If you have a 100k controller, and it fails, you need to think how to fix it. If your controller costs 30 bucks, throw it away and change for a new one.

Do they have spare batteries? Have the batteries been stored in an environment where they won't degrade? Are they rechargeable? If so, has their state of charge been confirmed before sailing, etc., etc., etc.

COTS stuff is awesome, but it doesn't absolve you from having proper procedures in place and knowing what those procedures should be in the first place.

but does a controller fail recoverably? Does the computer recieve old input (up/down) forever untill new controller is pligged in? how long does pairong proceas take? What if it fails at a critical moment?

Do you _know_ they had spare controllers?

Yes.

In a 2022 interview with CBC, Rush added the Bluetooth game controllers were durable — “it’s meant for a 16 year old to throw it around,” he said, tossing the controller to the floor — and that they kept spares on board “just in case.”

https://www.thestar.com/news/canada/2023/06/20/video-game-co...

Also, the controllers appear to be only to control the periscope. Which is not a critical component of this submarine. Everybody is fixating on them, but that's not what caused them to sink/get lost. It was probably the rest of the shoddy engineering.

Having spares is one thing. The question is do they test spares on a regular basis, make sure they have been charged between every mission and rotate them to test battery life?

No, controller on Titan is for movement as he talked about it (they don't have a periscope).

But he also mentioned a touch screen when talking about the controller so I bet that they can use the touchscreen to control the submarine as well.

Still most likely cause of failure is that the viewport failed and they died immediately.

US UAV/Drones use xbox controllers too

There was video floating around of a machine gun turret being remote controlled using the Valve Steamdeck in Ukraine.

https://www.tomshardware.com/news/steam-deck-controls-a-real...

Edit: consumer joysticks normally use potentiometers, which aren't great for deadzones/drift. For things like dust incursion reasons along it would make sense for any industrial/military device to be using hall effect based joysticks.

The more modern ones use hall sensor based sticks. Most new RC transmitter designs have 'em.

Valve donated few of those for kids. Most likely one of those units, than purposefully chosen.

Unmanned, if anything a controller failing will save some lives.

Unmanned, and they have logic to autopilot in most cases.

The periscope is a combat critical device, lose control of it and the enemy will see you first and you're dead.

Periscopes haven't been combat critical on submarines since slightly after WW2. They rely mostly on sonar to detect enemies, not vision - and of course they would. Periscopes are useless against submarines, and if an anti-submarine ship is nearby, you wouldn't go to periscope depth putting the submarine in a perilous position, and showing it off at that.

It's something that can be quickly swapped out if it does fail though being a wired controller, I'd put decent odds on this company not bothering to put a backup controller in their death tube. Also a periscope is less critical to combat in the age of sonar that can tell you bearing, heading and what type of ship often without the risk of surfacing and getting lit up on radar. Modern subs basically never want to surface in combat there's no need to take the added risk.

> this company not bothering to put a backup controller in their death tube

the really strange part of that, is that the pilot was the CEO of the company. Like the Norfolk Southern CEO would never in a million years set foot on one of their trains of death.

Anyhow they now heard sounds in 30minute intervals, so looks like they are still alive down there.

I saw the same thing but some of the groups searching haven't heard that pounding since Monday from the places I read. Does imply it might not have been a catastrophic implosion but honestly that's one of the better ways to go in a submarine. Also there's a lot of noise in the ocean on some other sub rescues they thought they heard noises from the crashed sub but it was just from the boats looking for them.

Quick replace is a fair point. Sonar completely superseding periscope is not quite as sonic countermeasures have been in use for decades. Also periscope depth is not surfacing.

Reply on news.ycombinator.com