4 ms·

The article mentions stair climbing as a requirement, not sure I buy that. Even wheeled robots can climb stairs with the right platform, I think that's still c

by fhd2 2y ago

The article mentions stair climbing as a requirement, not sure I buy that. Even wheeled robots can climb stairs with the right platform, I think that's still cheaper, less complex and error prone. But from a pragmatic perspective, it sounds like a very reasonable tradeoff to simply not support stair climbing. It's not like that thing is gonna walk itself from the workshop to the target area. If someone has to carry it anyway, they could conceivably carry it down some stairs while at it. (I'm a rather incompetent hobbyist when it comes to robotics, but I've been researching locomotion quite a bit and find a wheeled platform to be a good choice even in the forest environments I focus on. Guess it's just not exciting enough.)

kleiba 2y ago
The article mentions stair climbing as a requirement, not sure I buy that. Agreed. That requirement seems to be at odds with the premise of cleaning up beaches.
cowsandmilk 2y ago
1. Stair climbing isn’t just about movement from top to bottom. Stairs are a common place where people drop litter. Cleaning litter on the stairs isn’t accomplished by carrying the robot down the stairs. 2. Why do you believe the robot will always have a minder? The objective very much would be setting these off from a central location and covering a whole neighborhood or city.
- fhd2 2y ago
  If the point is to clean beaches and that's it, I think my points hold. Sure, there are stairs on some beaches, but it's not really any significant percentage of the space to be cleaned up, thus not really worth optimising for IMHO. If the point is to _start_ with beaches and to then use the same platform to clean up all kinds of yet to be determined areas, a quadruped might indeed be one of the few viable options. Wheeled robots can climb stairs with the right platform, but I wouldn't argue they can traverse arbitrary terrain the way quadrupeds can. Personally, when designing robots, I have very clear tasks and constraints in mind, making conscious tradeoffs. But I'm a software developer, that's how we do things. It's perhaps not how professional roboticists do things.

by fhd2 2y ago

kleiba 2y ago

The article mentions stair climbing as a requirement, not sure I buy that. Agreed. That requirement seems to be at odds with the premise of cleaning up beaches.

cowsandmilk 2y ago

1. Stair climbing isn’t just about movement from top to bottom. Stairs are a common place where people drop litter. Cleaning litter on the stairs isn’t accomplished by carrying the robot down the stairs. 2. Why do you believe the robot will always have a minder? The objective very much would be setting these off from a central location and covering a whole neighborhood or city.

fhd2 2y ago
If the point is to clean beaches and that's it, I think my points hold. Sure, there are stairs on some beaches, but it's not really any significant percentage of the space to be cleaned up, thus not really worth optimising for IMHO. If the point is to _start_ with beaches and to then use the same platform to clean up all kinds of yet to be determined areas, a quadruped might indeed be one of the few viable options. Wheeled robots can climb stairs with the right platform, but I wouldn't argue they can traverse arbitrary terrain the way quadrupeds can. Personally, when designing robots, I have very clear tasks and constraints in mind, making conscious tradeoffs. But I'm a software developer, that's how we do things. It's perhaps not how professional roboticists do things.