Wouldn’t a better solution be to mount several cameras and IR cameras strategically positioned just connected to a PC and an alarm? Maybe have a spotlight that can be directed at the location of the distressed swimmer so a real lifeguard can do the rescue.
It is hard enough to make a humanoid robot. Are you really suggesting making one that is water proof and can swim fast as well and get a gasping person out of the pool?
IMHO…NOPE.
Mike
Yeah, I don’t see that from the video. They’ve got one of those robots moving a single tote bin a couple of feet for the camera. Looks far more like they’re seeing how it works in the field, rather than “implementing” them as part of their process like all their other robots.
Plus the robot seems to be moving vastly slower, and more tentatively, than any human would - so I expect the speed will have to be much faster before they could implement a robot in any process in their warehouse.
The amount of processing power and the dynamic design of such an otherwise “low skill human” would seem to be exceedingly cost prohibitive. I would imagine a robot that could do all that would cost hundreds of thousands of dollars. The develop cost of such would be in the billions.
Again, an otherwise “low skill human” (yes, I know it isn’t no skill but lifeguards can be minors as young as 15 with less than a week’s worth of training) being replaced with a very expensive (water and weather proof), dynamic robot doesn’t seem to make financial sense. What community pool could afford the massive cost and maintenance such a robot would require?
So how many robots would they need to show that it is implemented?
Vastly how fast would that be? I think you need to be more specific. What you are stating is very subjective. I think the Robot is to short because they need to be at least 6 feet tall, but in reality is that even a valid point?
Andy
The word “implementing” implies there is a plan in process of being completed. If Amazon intends to roll out humanoid robots in their FCs I’d agree that is implementing. If they testing a few dozen that doesn’t sound like rolling out robots in the FCs.
Ok I will agree with that Syke. So if Amazon intends to roll out humanoid robots in their FC that is implementing. Let’s see what they say.
Andy
Never mind swimming and getting.
You would need a robot that could convince a drowning person to minimally cooperate so that victim and rescuer do not both stop “functioning” from water intake. On one lifeguarding rescue I did I actually had to tread water a few feet away from the idiot huge drunken drowning person, until he was so collapsed he could no longer try to climb on top of and drown me. A key part of my training was learning patience…
d fb
More than one. More than a bare handful.
Fast enough that replacing a human with that robot won’t result in a material slowdown in speed. If robots can perform human tasks, but only much more slowly than a human would, they’re not especially likely to be widely utilized. Especially where they’re part of a process, so that other parts of the business depend on their throughput.
I can recognize that a three-year-old is a child and a fifty-year-old is an adult, without getting bogged down in trying to figure out if the dividing line is 13 or 15 or 17 years old. A single robot is a pilot program, not implementation; a robot that’s taking 10x longer to carry a bin four feet than a human would is too slow. So seeing a single video of a robot slowly carrying a bin for a camera isn’t much evidence that Amazon is actually implementing these things. Though, again, it’s great content if you’re trying to get people excited about the prospect of humanoid robots.
In our training we learned to dive below the surface about 10 feet from the victim, grab his legs and turn him so he faced away from us, then surface and strap our arm across his chest keeping his head out of water by nestling it on our shoulder.
Comment to keep the post on-topic: Yeah, water-borne automated humanoid lifeguard robots seems a bridge too far, right up there with helicopter cars or light-beer that tastes good.
Why would it have to swim at all? Let it roll on the bottom (waterproof, duh) and use some type of lift platform to push someone above the surface of the water.
Heh.
A very non-human form that would probably work in confined spaces like a pool.
That being stated, we would still either need a human to identify people struggling from the surface or develop very expensive AI software and cameras that can differentiate someone drowning from someone simply floating or just splashing water.
dlbuffy wrote:
Jetson-thor, Groot, etc…WOW…how much are those royalties going to cost? LOL. I actually wish they weren’t trying to be cute, makes me sus.
This is actually a key part of the Disney design. Those robots are for entertainment and are specifically designed to be childlike and mischievous.
You are right to be sus. They are not really a worker, they are performing artists.
eta: I think you may have been talking about the naming conventions for their process/tech stack. This is excellent advertising and is extremely relatable for humans. (just like the Apple iOS names and the AndroidOS, and, and, and…)
Why ever would I expect that a ‘lifeguard’ robot would be humanoid?
Several folks are DEMANDING that ALL factory floor robots MUST be on wheels and shaped like a tank vacuum cleaner.
Personally, I think the form factor will be ‘appropriate’ to the situation.
Shouldn’t I apply that same logic - ie the APPROPRIATE form factor for the job?
Therefore, I propose the lifeguard robot will be crocodilloid.
Or, perhaps Jawsoid.
Although perhaps someone with less imagineer in em, might envision a bot with some kinda ‘scoop-net’ appendage. Oh, and floatation devices.
The REAL need is an AI that uses the bot’s abilities to ‘stop the drowning, administer first aid as appropriate to ‘water environment’, move the victim to shore or rescue boat, so other bots can do THEIR job’… etc.
ralph swims poorly. slowly drifts downward…
but with Mask,fins,snorkel… get outta my way!
No one’s demanding that. We’re just pointing out that there’s rarely any situations for factory floor robots to have legs, or have differentiated heads and torsos shaped exactly like human ones. Which makes the decision to design robots with those features more an aesthetic choice, rather than a practical one.
Suppose a hotel or YMCA wants to get the most bang for the robot buck. So in addition to lifeguarding, the robot will also put the deck chairs away, pick up trash, brush down the pool, and clean the locker room and showers.
What form factor is best suited for a bot that can do all those activities? I don’t know the answer, but I do know that a human being is capable of doing all those tasks.
So as I see it, the business could buy several specialized robots, each with form factors optimized for each task. Alternatively it could buy a single humanoid robot capable of using tools made for humans to do all those tasks.
I’m guessing the second option is more likely to be cost-effective.
Ok so when I show you a picture of 7 humanoid robots working in a factory you will agree that it is implemented. That sounds fair enough.
Andy
Elon Musk has explained this. Jensen Huang just explained this. Whether you choose to understand or agree with it is completely up to you. I choose the experts.
Andy
Depends on the factory. If the factory only has a dozen workers, and seven are humanoid robots, absolutely. If the factory has three thousand workers, and there’s seven humanoid robots working in a pilot test of those robots, it wouldn’t be implementation.
I say that because it’s very likely that at least one of these companies will make a tranche of robots and partner up with a company to test them out - and there very well may be seven in that test program. As with any Hot New Thing, for some of these startup the stock is the product - that’s what they’re selling to get rich - and an easy way to promote the stock is to get some videos of robots working in factories. Whether they’re useful, important, or valuable to the factory - they’re incredibly useful to the company that wants investors or financing.
Just like with blockchain, there’s going to be a ton of hype to try to part people with their money to invest in these efforts. Whether it makes any sense or not.
LOL you said more than a handful. Now you are giving qualifiers to your qualifier. Brilliant.
Andy
They haven’t explained it. They’ve handwaved it, saying that robots need to be shaped like humans in order to work in human places. But that’s not true. Our world is filled with robots, and none of them (yet) are shaped like humans. They haven’t yet explained why a robot with any brain that we can build in the next decade or so needs to be shaped like a human.