Developing an underwater robot to save people from drowning • Hacker News Board

Re: Developing an underwater robot to save people from drowning

#11

Post by meheleventyone » Tue, Mar 09, 2021, 11:55 AM UTC

jmercouris wrote: post #2

If it is using computer vision to detect a drowning person, why not just signal an alarm so people can help?

Yup, a poor chap died in my local pool a few weeks ago despite lifeguards on duty and cameras underwater in the pool. It took a member of the public spotting him underwater to alert anyone by which time he'd been under for six minutes and was dead. Literally a couple of meters away from the lifeguard with another in a "command center". Pool drownings are exceptionally rare but it's terrifying to think how much the ball was dropped in that case.

Re: Developing an underwater robot to save people from drowning

#12

Post by frobozz » Tue, Mar 09, 2021, 12:03 PM UTC

beckman466 wrote: post #8

> "During testing at the Hufeisensee lake, the dummy did not slip out of the securing mechanism as it was taken to the surface." > A mechanism for fixing the rescuee in place prevents lifeless bodies from sliding down as they surface. That looks and sounds like a deathtrap. Does it really try to strap the person to a hospital bed, underwater...? Won't that drown the person if the robot is too slow or malfunctions? >…

> Does it really try to strap the person to a hospital bed, underwater...?

It looks like it folds two barriers against the sides of casualty to pinch them in place. I imagine that if you are not completely incapacitated, you can dodge or wriggle out of it. I would also guess that, given that it's supposed to be clever enough to spot an incapacitated casualty to rescue, it might be clever enough to stop trying to rescue you if you attempt to swim away.

> Won't that drown the person if the robot is too slow or malfunctions?

WRT rescuing unconscious casualties from dangerous situations, you start with the assumption that they will die if you do nothing, and could already be dead. With that in mind, who cares if this robot drowns someone halfway to the surface who would otherwise have drowned at the bottom?

> So the reasoning is: job training for lifeguards is not feasible (why?), so we should develop super expensive robots that local pools can't afford.

I agree with you on this point. I think there is something in having a system that can help spot swimmers in danger, but then leave the job of rescuing to actual lifeguards.

The problem of not having enough lifeguards is not that they cannot perform rescues, but that they cannot continuously observe the entire area and spot people in need of help. Apart from spinal injuries, a single lifeguard is perfectly capable of performing a rescue in most pools or swimming ponds (in much larger or more complex areas, like water parks or the sea, it might be harder to get to a distant casualty from any given station).

Another thing is that lifeguards prevent the situation that this robot is designed to rescue people from. You spot a swimmer in trouble, and toss them a float or a line. You spot a dangerous situation arising, and you take action to prevent it. You don't wait for them to sink to the bottom and pass out.

Re: Developing an underwater robot to save people from drowning

#13

Post by dopidop » Tue, Mar 09, 2021, 12:17 PM UTC

anonymousDan wrote: post #9

qchris wrote: post #5

I think this kind of a system is extremely unlikely to be applied in practice, at least in the United States--it's just too costly. You can reasonably train a 16-year-old to be a lifeguard in less than 30 hours[0]. Meanwhile, even the most basic of prosumer ROVs (remotely-operated underwater vehicles) like the BlueROV2 costs almost $3000[1], that that's without including the fancy computer vision tech and/or acoustic…

You are mixing up a one off cost (buying the robot Vs training a lifeguard) with the most likely much more significant ongoing cost of the lifeguard wages.

So, 8$ an hour?

Re: Developing an underwater robot to save people from drowning

#14

Post by deepstack » Tue, Mar 09, 2021, 12:20 PM UTC

beckman466 wrote: post #8

> "During testing at the Hufeisensee lake, the dummy did not slip out of the securing mechanism as it was taken to the surface." > A mechanism for fixing the rescuee in place prevents lifeless bodies from sliding down as they surface. That looks and sounds like a deathtrap. Does it really try to strap the person to a hospital bed, underwater...? Won't that drown the person if the robot is too slow or malfunctions? >…

So the reasoning is: job training for lifeguards is not feasible (why?), so we should develop super expensive robots that local pools can't afford.

Yeah when I first read this article. I was thinking "Lifesaver". Cheaper and just as effective.

Re: Developing an underwater robot to save people from drowning

#15

Post by anonymousDan » Tue, Mar 09, 2021, 1:20 PM UTC

dopidop wrote: post #13

anonymousDan wrote: post #9

Earlier quoted context omitted.

You are mixing up a one off cost (buying the robot Vs training a lifeguard) with the most likely much more significant ongoing cost of the lifeguard wages.

So, 8$ an hour?

Sure, 8 times 40 hours a week times 48 weeks a year is roughly 15k.

Re: Developing an underwater robot to save people from drowning

#16

Post by LeifCarrotson » Tue, Mar 09, 2021, 1:59 PM UTC

meheleventyone wrote: post #11

jmercouris wrote: post #2

If it is using computer vision to detect a drowning person, why not just signal an alarm so people can help?

Yup, a poor chap died in my local pool a few weeks ago despite lifeguards on duty and cameras underwater in the pool. It took a member of the public spotting him underwater to alert anyone by which time he'd been under for six minutes and was dead. Literally a couple of meters away from the lifeguard with another in a "command center". Pool drownings are exceptionally rare but it's terrifying to think how much the ba…

Here, try not to drop the ball yourself:

http://spotthedrowningchild.com/

And remember that as you said, pool drownings are exceptionally rare, so instead of knowing there's a drowning child in frame in the first 60 seconds or so, you're watching for 8 hours a day and probably won't see one all summer.

As an industrial controls engineer, I find myself constantly explaining that humans are pretty bad at being robots and robots are pretty bad at being humans. This seems like yet another example. People are really bad at paying extremely close attention to their entire field of view continuously; machine vision does that tirelessly for the price of a few watts of electricity. In this case, the underwater robot is helpless to communicate to other poolgoers and to the suspected drowning victim, and as a mass of propellers and PVC it's probably slower than a skilled swimmer. Others have expressed concern about being trapped in a potentially broken robot's grasp: robot actuators are bad at grasping squishy humans, but humans are good at that.

I think the greatest value in this system is in the cameras. A light and/or sound to notify a nearby human would make lifeguarding hugely more effective and easier. That said, I don't particularly look forward to camera-laden advertising zepplins floating over the local swimming hole.

Re: Developing an underwater robot to save people from drowning

#17

Post by melomal » Tue, Mar 09, 2021, 2:21 PM UTC

beckman466 wrote: post #8

> "During testing at the Hufeisensee lake, the dummy did not slip out of the securing mechanism as it was taken to the surface." > A mechanism for fixing the rescuee in place prevents lifeless bodies from sliding down as they surface. That looks and sounds like a deathtrap. Does it really try to strap the person to a hospital bed, underwater...? Won't that drown the person if the robot is too slow or malfunctions? >…

> job training for lifeguards is not feasible (why?)

Growing up anyone that could swim well was desperate to become a life guard. Well above minimum pay and for 16+ that's pretty good. Plus there's an aura about being a life guard that was incredibly interesting. You get to wear shorts and t-shirts all day too.

That being said I assume most of these upcoming life guards have disappeared to do a tiktok dance...

Re: Developing an underwater robot to save people from drowning

#18

Post by craftinator » Tue, Mar 09, 2021, 2:51 PM UTC

frobozz wrote: post #12

beckman466 wrote: post #8

> "During testing at the Hufeisensee lake, the dummy did not slip out of the securing mechanism as it was taken to the surface." > A mechanism for fixing the rescuee in place prevents lifeless bodies from sliding down as they surface. That looks and sounds like a deathtrap. Does it really try to strap the person to a hospital bed, underwater...? Won't that drown the person if the robot is too slow or malfunctions? >…

> Does it really try to strap the person to a hospital bed, underwater...? It looks like it folds two barriers against the sides of casualty to pinch them in place. I imagine that if you are not completely incapacitated, you can dodge or wriggle out of it. I would also guess that, given that it's supposed to be clever enough to spot an incapacitated casualty to rescue, it might be clever enough to stop trying to resc…

> I would also guess that, given that it's supposed to be clever enough to spot an incapacitated casualty to rescue, it might be clever enough to stop trying to rescue you if you attempt to swim away.

One would hope, but I wouldn't hazard a guess at that. It likely uses some type of ML/NN to spot an incapacitated person; there are a bazillion pretrained models that recognized people in images. I know of no models that recognize when someone is "trying to escape". I think that that would be quite a difficult problem to solve, except for possibly a number of panic buttons. But in general, guessing a person's intent programmatically is really hard.

Re: Developing an underwater robot to save people from drowning

#19

Post by klmadfejno » Tue, Mar 09, 2021, 3:17 PM UTC

frobozz wrote: post #12

beckman466 wrote: post #8

> "During testing at the Hufeisensee lake, the dummy did not slip out of the securing mechanism as it was taken to the surface." > A mechanism for fixing the rescuee in place prevents lifeless bodies from sliding down as they surface. That looks and sounds like a deathtrap. Does it really try to strap the person to a hospital bed, underwater...? Won't that drown the person if the robot is too slow or malfunctions? >…

> Does it really try to strap the person to a hospital bed, underwater...? It looks like it folds two barriers against the sides of casualty to pinch them in place. I imagine that if you are not completely incapacitated, you can dodge or wriggle out of it. I would also guess that, given that it's supposed to be clever enough to spot an incapacitated casualty to rescue, it might be clever enough to stop trying to resc…

I give it 7 seconds before the children start pretending to drown at the bottom of the pool to summon the rescuebot.

Re: Developing an underwater robot to save people from drowning

#20

Post by klmadfejno » Tue, Mar 09, 2021, 3:22 PM UTC

LeifCarrotson wrote: post #16

meheleventyone wrote: post #11

Earlier quoted context omitted.

Yup, a poor chap died in my local pool a few weeks ago despite lifeguards on duty and cameras underwater in the pool. It took a member of the public spotting him underwater to alert anyone by which time he'd been under for six minutes and was dead. Literally a couple of meters away from the lifeguard with another in a "command center". Pool drownings are exceptionally rare but it's terrifying to think how much the ba…

Here, try not to drop the ball yourself: http://spotthedrowningchild.com/ And remember that as you said, pool drownings are exceptionally rare, so instead of knowing there's a drowning child in frame in the first 60 seconds or so, you're watching for 8 hours a day and probably won't see one all summer. As an industrial controls engineer, I find myself constantly explaining that humans are pretty bad at being robots a…

I used to be a lifeguard. My pools weren't nearly as busy as that website's examples. If they were, I probably wouldn't have done the job, as that looks super stressful. Much to my surprise, however, I did spot the drowning child at around the same time as the lifeguard. Youth employment validated.