[–] Xantha 0 points 3 points (+3|-0) ago 

Automotive production engineer here, who happens to deal with fasteners in particular.

You can get a robot to mount and unmount bolts or nuts all day long correctly as long as A. The robot is running correctly B. The mounting fixture has no defects present C. The fastener (nut or bolt) is in within required specifications.

A decent number (not all) fasteners are hand-started, then the production line operators torque it down with a torque gun (or a nut runner). I don't think I've ever seen someone use a hand-wrench (of any kind) to tighten something on a vehicle. (Because it's wildy inconsistent between people, robots are lightyears more consistent--- I know completely contrary to what a bunch of people say in a lab in this article)

Torque guns use exact limits to verify that fasteners are tightened correctly because it's a major safety issue. You can bet your ass if the torque gun stops/faults out .01 out of spec, the OEM (Nissan for instance) will throw a fit over it and contact the supplier to ask why the parts were fucked up.

The reason there aren't robot arms (outside of welding and lifting) on every major vehicle production line fastening everything instead of people is A. Cost B. Cost C. The fucky-ass designs the design engineers come up with that make it really hard to mount shit without using a person.

Toyota famously (infamously) loves to tell a story about how one of their line operators was having trouble mounting some fixture in vehicles he was working on. He asked them to design a chair based off his fly fishing boat chair (that can slide back and forth) that made installation significantly faster and more reliable. Now every Toyota plant in the world has these chairs for that particular part of fixture installation.

Was the chair necessary? No, had the vehicle and/or the fixture been designed a different way, the chair probably wouldn't be needed at all. But once you get into calculating (just) the labor to redesign a whole section of the vehicle, accounting goes "No, no, fuck off, do something else with your time."

I would love to see how these group plans to make disassembly (for any product) cost effective enough to use robots on. Individual components are (almost) always worth less than the entire assembled product. Are they going to resell the disassembled components as (new/used?), are they scrapping the disassembled components (for fractions of a penny on the dollar)?

If there were any precious metals (valuable enough) in a smartphone (for instance), I have a hard time seeing how anyone would invest big $$$ into disassembly when you could incinerate/liquify the entire product and chemically separate the item you're looking for in a vat.


[–] B3bomber 0 points 0 points (+0|-0) ago 

That is the biggest cost of designing shit for robots. Designing the shit for robot interaction when you no longer need a human. This requires the process be designed for robots from start to finish and there is no chance they will pay for that R&D.

Anyone who figures out how to do that swap cheaply and quickly will completely fuck every other manufacturer once it's done.