Industrial robots, such as those used by automakers, make it possible to automate many repetitive tasks that would require many hours of manual labor, but is it efficient?
Toyota Motor Corporation found out the hard way that industrial automation is not as efficient as having real human beings doing the work. In the short run, it may make sense to have machines stamp out, glue, and weld parts together, hundreds at a time. Toyota is the world’s biggest automaker, and automation helped it to get there. On the other hand, perhaps there is such a thing as excessively rapid growth, and overreliance on the machines has led to complacency on the part of the company.
Case in point, certain Toyota vehicles, in heavy rains, would allow water to enter the passenger compartment by the driver’s feet. The problem was a lack of sealant in the area where the firewall meets the side panel. During assembly, just before positioning these parts for welding, a robot precisely draws a line of sealant where the parts are supposed to meet. At some point, there is a zig in this line of sealant where it should have zagged, resulting in an unsealed area. What could have been solved with a tweak of the sealing robot’s programming later costs hundreds or thousands of dollars in leak repair and mold remediation. This is not very efficient.
Toyota is going back to manual labor in some parts of the automobile-manufacturing process, and the results are important to note. For example, at Toyota’s Honsha plant, master craftsmen, instead of machines, are forging and cutting crankshafts. Leaving the robots out of it has actually been more efficient. The production process, for crankshafts alone, produces 10% less waste than before, and Toyota has been training and retraining their master craftsmen on every part of the manufacturing process. These, in turn, replace machines doing these tasks and, at time, work alongside the machines that do the more repetitive tasks, monitoring and tweaking their function for maximum efficiency and quality.
Photo credit: Aaron Webb / Foter / Creative Commons Attribution-NonCommercial-ShareAlike 2.0 Generic (CC BY-NC-SA 2.0)
NeilFarbstein unfortunately, most human interaction in manufacturing these days consists of pushing something into a machine and calling someone else when there is a problem. cheap labor doesn’t need to know any more than that. it saves money in the short run, but we lose out in the long run. need to get the craftsmen and professionals back into the mix.
besides, we’re far from true ai robots that teach themselves. still need a master to teach the machine.
They should keep human beings with those skills around to supervise the machines.