This shit is scary. People don't know what's going to hit them.
(media.scored.co)
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (165)
sorted by:
The "simulation" part doesn't matter. It won't help you actually use robotics to repair themselves in the real world.
I can't quite fathom the confusion of thinking that would cause you to make this comment.
It is exactly like pointing to the steam train and saying "Well, it doesn't do naval transport!."
So what?
If you can build a knowable robotics assembly cell. That is, a cell with a robot tool in it, for which the calibration, control systems, physical distances etc are known. Then you can put that information into a simulator and then use it to grade the responses of Deep Learning Neural Networks.
Yeah, some robotics tools will be better than others, and some will be more capable than others. You can simulate that too! Now you have the potential to test a million robotic work cells, and do it all before you spend your first dollar on physical robots. You can even optimize designing robots to build other, better robots and bootstrap a tool chain.
Suddenly assembling the parts of a robot bunny becomes a software problem. It is scalable and measurable. You can recruit computer network arrays to simulate the work and grade it, then give that feedback to the NN.
Now complex problems of robot manufacturing can be approached with exactly the same tools that is having Deep Learning Neural Networks paint like Rembrandt or write novels like Joke Rowling.
Look, it does not matter if you believe me or not. It is coming down the pipe. Just sit back and relax.
So you're trying to do naval transport, and handing me a blueprint of a steam train.
Except how you don't. Blah blah blah Neural Network, still non real nonsense. Absolutely relevant to jack shit. It cannot be used in the real world without actually doing the real world work. You can't get around that hurdle, anything else is just jerking off. It is not and never will be solely a "software problem" because at some point you actually have to cross the line into physically doing something.
So without fapping on about non real bullshit, answer the fuckin question already.
How do you propose to actually DO anything at all to create a robot capable of repairing itself in the real world, without a human doing said work physically?
Are you saying that it isn't possible to measure and then simulate a robot doing sophisticated manufacturing work?
Here is a video of a robot following a program that was created by algorithm and then tested in simulation, before manufacturing a real, usable product.
https://youtu.be/CL-I-RvMeAU
You will notice that it is fourteen years old.
Autodesk makes Fusion 360 a product specifically designed and built to run CNC machines, which includes accurate simulators. There are tens of thousands of videos of both the simulation and then the actual cutting of parts. Go and look!
Now why do you think that this process can't be used to grade Neural Network output?
Is it that you don't know that CNC tools are robots? Or that you think that the same techniques can't be applied to a six axis robot arm with a tool? Is that your issue?
https://youtu.be/Hze1D_B2ui4
Here is a three year old video of a robot arm using cold spray additive manufacturing to print a solid copper rocket nozzle bell housing. The tool paths were generated by algorithm and then simulated before the print.
Where do you think I said that robots would be able to repair themselves? Can you highlight that for me?
Why would replacing technicians that do repairs on equipment be more important than actually NN tools to reduce the cost of the design, manufacturing and assembly of parts?
It really seems like you are responding to an argument I didn't make.
Because that's the entire conversation. Whether robots will replace tradesman whose job is robotics. That is the entire topic.