Hacker News Reader

Comment by ogre_codes | Hacker News Reader

So the problem they seem to be solving is slowly moving small things short distances across really rough terrain with a power source nearby. I don't see these flying off the shelves.

The bulk of everything else it does could probably be done better with a drone. The only big win for the dog-bot is the 31 pound carry capacity but... I'm not even sure how that is super useful.

View on HN

Indoor environments where a flying creature is a real hazard, especially since you can't have a drone above people (typically) but you can have a dog robot walking around as long as it's not so slow that it blocks stairs/doorways.

Spot with no load (or minimal load) seems to be pretty fast? I could imagine it could do useful checks around a construction area or an empty building and alert someone if something is off - "Hey, operator, a door was opened and now it's closed."

Obviously it's early tech, we're going to see first wave adopters who have an appetite for experimentation.

Wheeled robots already dominate indoor environments. They cost a tiny fraction of what this costs, have better endurance, and are safer.

Stairs, debris, uneven walkways. Spot does not fall even in really hazardous environments and I have seen some aggressive torture tests.

[Youtube warning] https://www.youtube.com/watch?v=R-PdPtqw78k

Much like drones are much better at most outdoor operations, wheeled robots are much better at most indoor operations. Yes, stairs, etc etc... but if a company has a big enough need for investing in robotics, it's likely they will be willing to invest in the infrastructure (ramps, etc) to support those robots.

Also, consider for a moment the following from Boston Dynamics:

> Spot should always be operated at least two meters away from people

That alone eliminates something like 95% of workplaces.

> That alone eliminates something like 95% of workplaces.

Not currently, as people also operate at least two meters away from people.

Flying drones have major endurance issues, especially if they need to carry anything beyond an extra battery. This can last 90 minutes on a 4.2 kg battery and presumably extend that to 4.5 hours at the cost of ~60% carrying capacity. Drones also have trouble applying forces do do stuff like open doors.

75k seems like a much bugger issue in terms of general use, but that can probably drop a lot over time.

> Flying drones have major endurance issues

What's the point of lasting 3 times longer if you are moving 5-10 times slower? Even if you double the endurance of the Spot, you are still covering the same amount of ground that a much more affordable drone could cover.

Speed is great have a specific well defined destination, but that's not always the case. Consider using a spot vs a drone to walk your dog.

You got me there. I'm sure this thing will totally corner the market on $75,000 dog walking robots.

At 75,000$ using a person is almost always going to be a better solution right now. Even above 20k I suspect it’s mostly just people doing R&D. However, that is kind of like looking at the first version of the segway and ignoring the cheap ”hoveboards” that showed up at 1/60th the price. At say 1,000$ you can bet some people would use them for dog walking.

I should know better than use sarcasm here.

> However, that is kind of like looking at the first version of the segway and ignoring the cheap ”hoveboards” that showed up at 1/60th the price. At say 1,000$ you can bet some people would use them for dog walking.

The Segway was from the get go a product with immediate and obvious uses. It didn't take off the way Kamen thought it would, but it was a product. This by contrast is barely a product at all, it's almost 100% a research tool. Even most of the people who use this thread suggest as much.

I think it's still best suited to be a good research vehicle for interesting AI work.

I’m pretty sure the main use is observing dangerous environments like mines, places with toxic waste, and oil rigs. The payload will usually be used for sensors or a door opening device, not to transport cargo.

> observing dangerous environments like mines, places with toxic waste, and oil rigs.

This could be a good use. Mines in particular where drones would be impractical. I'm not so sure about oil rigs, while Spot would be great on the stairs/ terrain, it has trouble with seeing the edge of cliffs and might well march right off into the ocean.

Probably the biggest advantage would be the ability to take samples or hold some kind of special equipment.

In fact, there is a startup called Exyn Technologies which tackles the problem of deploying autonomous drones in mines and other such dangerous environments (https://www.exyn.com/)!

Reply on news.ycombinator.com