Comment by renewiltord | Hacker News Reader

Comment by renewiltord | Hacker News Reader

So what's the deal with these guys? Obviously their tech is pretty cool since I remember the control and feedback problem for bipedal bots being described as really hard when I was growing up. Obviously their consumer marketing is great since everyone looks at these things and loves them.

But they don't have a product. What's the problem? My guesses:

* No one really needs this shit

* They have a control/feedback problem in interaction with moving objects i.e. this thing can't high-five you without breaking your hand or open a can

We usually handwave this shit with "But The Military" but like what's the use-case for the military? This fucker doesn't know how to react to anything so it's a bipedal trolley. But it's too expensive a trolley. If you take this guy with you, your mission is going to end up being to protect this guy. Take an extra human and he's way cheaper, way more versatile, and all that.

Not trying to be the usual HN negative because I actually think this is like the Newton or something. The tech is freaking cool, and it excites me as a glimpse to the future, but the product part is going to come some time in the future and it's going to be awesome and in hindsight we'll be like "Damn! That was the missing part. It's so obvious". Right now, though? I just hope someone will keep buying them and selling them like a hot potato so they live.

They are still trying to figure out their market. They are essentially selling a product the world isn’t quite ready for.

If it was dirt cheap I would buy one, attach a vacuum to it and have it clean my place. But they are super expensive, so fat chance of that.

Spot, the yellow dog, seems to be targeted at the notion of human safety. Think something like a building demo job gone wrong, send in the robot to pull out the dynamite.

In Norway they are being tested on oil platforms. If this robot can be used to fix one single problem on a platform that would otherwise have required a production shutdown...it’ll pay for itself several (hundred?) times over. This is especially the case for older platforms where you need to be more hands on and less can be done remotely.

We had a mountain side in Norway that was near collapsing and all the residents below it were evacuated every two weeks. The geologists wanted to use dynamite to dislodge it, but it was too dangerous for humans. In the future we might send a versatile robot like Spot up in a drone and have him plant the explosives. If he blows up in the process that would obviously suck, but still fairly cheap for what he would have accomplished.

Found a relevant link: https://akerbp.com/exploring-the-potential-of-robotics-in-th...

> We had a mountain side in Norway that was near collapsing and all the residents below it were evacuated every two weeks. The geologists wanted to use dynamite to dislodge it, but it was too dangerous for humans. In the future we might send a versatile robot like Spot up in a drone and have him plant the explosives. If he blows up in the process that would obviously suck, but still fairly cheap for what he would have accomplished.

Why not just drop a bomb from an airplane? That's already pretty well established technology. I'm pretty sure I've read about various places throwing hand grenades or dynamite from helicopters by hand to do that. Or let the nearest friendly air force drop a few of their lightest bombs on it for practice.

They tried dropping water during the winter and let it freeze.

I think dropping a bomb would do more harm then good, and might not even dislodge anything. I think they wanted to drill close to the fault line and insert explosives deep into the mountain.

The way I see it, they're just doing R&D. The things they show off serve as advertisements to attract new investment, and remind the current investors that they're making progress.

And it's fine, really. They push robotics forward, they solve really tough problems. It's infinitely more interesting and useful long-term than most of the high-valuation startups on the market these days. I'd take another BD for a different field over another DoorDash or WeWork any day.

If you’ve been following them since the first video, now that the time horizon is long enough, you can start to appreciate the value of iterative incremental improvements compounded through time.

One actual use-case of DARPA ("The Military") is to deploy such robots in disaster-stricken areas instead of risking human. For example, emergency maintenance in a power plant that has suffered a nuclear meltdown accident.

> But they don't have a product.

The yellow robot, Spot, is available for purchase.

Right, I know they run it in Singapore, but it's a product the way Juicero is a product: it's kinda useless.

They do have a product you can buy — Spot (the dog).

Currently a smooth 75,000. Alas I am left dreaming of my robo dog companion :(

Reply on news.ycombinator.com