2 ms·
Why is either of these a case for a humanoid robot, and not a specialized robot? Humanoid is a pretty awful design for any specific task; it's just that we've e
by danaris 6y ago
Why is either of these a case for a humanoid robot, and not a specialized robot? Humanoid is a pretty awful design for any specific task; it's just that we've evolved and learn to be good at many tasks.
In the warehouse, you can adjust the design of the warehouse a relatively small amount to ensure that the robots can easily store and retrieve the packages—which come in a finite number of standard sizes and shapes, which in turn can be adjusted to ensure they pack well while still leaving room for the robot's grasping arm/claw/whatever to grab it. Depending on the specific designs, it may be worthwhile to combine the store-and-retrieve robot with the carry-cart. It's certainly not going to be more effective to have a humanoid robot that has to climb up ladders, rather than some sort of wheeled or tank-treaded robot that can anchor itself to the floor and extend a telescoping part of it up to get or leave the package.
For picking strawberries, even if we accept that you want "hands with the same dexterity and nuance as a human", why would those hands do best attached to a human body? Strawberry plants are pretty low to the ground, so why not put them on a little rolling bot that can go between the rows right at berry-level?
No; the best case for a humanoid robot is in human-facing positions. We like to interact with people, and a robot (note, not a human-level AI: a robot; human-level AI gets into serious ethical and philosophical quandaries real fast) doesn't care if you yell and scream at it. It'll just keep saying, "Yes, sir. Please give me your account number so I can proceed," for hours on end if need be.