Hacker News Reader

Comment by perihelions | Hacker News Reader

- "...active control system..."

https://spectrum.ieee.org/this-maglev-heart-could-keep-cardi...

That's some high-stakes software programming!

Passively-stable magnetic levitation is possible with diamagnetic materials (notably including superconductors), but those probably don't meet the requirements of this project.

- "This positional control system works as follows: Tiny contactless sensors send out magnetic fields that interact with the rotor, determining its exact location many times per second. If the rotor is moving in one direction or another, the control system puts electrical energy into electromagnetic coils within several actuators, causing them to cancel out that movement."

View on HN

I know these aren't intended for long-term use, and presumably not in ambulatory patients, but still, this passage gave me a tremendous amount of range anxiety:

> Patients will wear a 4-kg external controller pack that contains two rechargeable batteries (providing about 5 hours of operation each), although they can also plug in directly to a power outlet.

Like, the stress of finding a free outlet when I'm on 7% charge is already bad for my heart

That's heavy battery package. Is the heart much more efficient than these devices? I'd doubt so.

Does anyone know if there's been research to extract energy from the blood directly?

given that the heart manages to operate inside the body without those two external battery packs, I'd wager that it's more efficient.

The heart has access to the resources of the body.

Does the human body provide more power to the fleshy heart than two large external batteries?

The heart is massively more efficient. Just about everything in all of biology is massively more efficient.

Take a moment to consider that your entire body runs on an energy budget of only around 100 watts, of which your brain only uses a measly 20 watts.

This writeup is significantly better than the OP link. Should probably be the article of record.

The section on constant flow vs. variable flow to simulate a heartbeat got me thinking about how weird it would feel to have no pulse, or what it’d be like to experience a moment of excitement/adrenaline without the accompanied pounding in my chest.

Yes, but not as high-stakes as an autopilot in a commercial aircraft!

Perhaps, but there are a pilot and first officer to take over if the autopilot goes out. You better be in the OR if your titanium heart tries to kill you...

Exactly. And systems for commercial aircraft are actually designed (by law) to be dual fault tolerant. This means that all safety critical systems on an airliner need to have 3 filly redundant, independent systems and each needs to be able to function without the other two.

In even class III (implantable and safety critical meaning the patient is harmed or killed if they fail) medical devices are only required to be single fault tolerant.

>And systems for commercial aircraft are actually designed (by law) to be dual fault tolerant.

Not true. Proof: Boeing 737MAX MCAS system, which relies on only one sensor, and when that sensor ices up, causes the aircraft to fly into the ground, killing many hundreds of people in two separate incidents. This design was fully approved by the FAA.

Yeah, that's a famous example of a failure in the process.

The deal there was the MCAS system was originally designed such that it was thought to be not safety critical. It relied on another sensor that was also non safety critical in its other uses. But then over time Boeing made changes that made it safety critical, and rather than upgrade the hardware they put some footnotes in the manual and expected pilots to do things to avert disaster, but then they didnt properly train pilots to those changes to their duties.

It was a comedy of errors on many levels.

None of it changes my parent comment.

Article doesn't explicitly say that, is there actually a software running inside this heart? That's very anxiety inducing.

What are the consequences of a rotor "crash"?

Reply on news.ycombinator.com