Foxconn Factories' Future: Fewer Humans, More Robots

jfruh writes: Foxconn, which supplies much of Apple's manufacturing muscle and has been criticized for various labor sins, is now moving to hire employees who won't complain because they're robots. The company expects 70 percent of its assembly line work to be robot-driven within three years.

Comments

Automation is Dependent on Design for Manufacture

[mtippett]

I've been to Foxconn factories in Shenzen, and there are clearly opportunities for deeper automation. However, this will only be possible when the underlying hardware design has been designed for automation.

At the PCB level, pick and place achieves amazing automation and performance with smaller than rice-grain size components used in modern electronics. That is a given.

At the assembly level it isn't so easy to automate with a lot of the designs. There are flex cables, adhesive, torque sensitive screws that all rely on a human to be able to manipulate and then quickly respond to misalignment. To automate this, the design constraints placed on the Industrial Designs need to change. For low and mid-range products where form is not at the level of Apple integration, this will probably increase the automation. For the high end where every mm counts it's unlikely that there will be a high level of assembly automation.

Re:Foxconn Factories' Future: Fewer Humans, More R

[khallow]

Right, but the conversation that's being had around this is what are we going to do with all these people that we don't need anymore.

Employ them. That's what China will do. The two or three "lost generations" is a developed world problem coming from an uncompetitive labor force.

Re:Automation is Dependent on Design for Manufactu

[Kjella]

At the assembly level it isn't so easy to automate with a lot of the designs. There are flex cables, adhesive, torque sensitive screws that all rely on a human to be able to manipulate and then quickly respond to misalignment. To automate this, the design constraints placed on the Industrial Designs need to change.

I think you underestimate how far sensor technology has come and will go, here for example is an example of automated salmon processing. Obviously there's a lot of natural variation, do we need to bioengineer a more robot-friendly salmon? No. They're measured out by a laser and intelligently cut. Head/tail/other cuts are dropped out to go on another processing line. Each cut is grabbed by a robot with robot vision and placed in pouches to be sealed. Skip to 3:12 if you just want to see that last part. Fillet-making machines are still in the research phase but there are examples of that too using X-rays to scan and find the pin bones. If they can deal with all that, I'm sure they can apply the right torque to a screw.