Let me be clear, if you spend circa US$100k you can get a cool six axis robot with around a meter of reach, 10/15 kilos of load capacity, 0.030 mm precision in any axis, a camera vision system and a suitable end tool / gripper.
At this point you are ready to “train” it to do one repetitive task, and pretty much, the task must come to the robot.
That “task” can be picking up shit from a container or palette on the left, placing it in front of another machine tool in front while something is done to that part, and then placing the finished part into a container or palette on the right.
You’ll need someone to place and remove the palettes, and set it all up and train everything.
Change the part but not the process, more training and possibly another gripper.
But, it will never get tired, never sue you for sexual harassment or discrimination, never go sick or get pregnant, never go home or work for someone else.
It will however have ongoing costs, wear and tear an calibration, an electric and air / vacuum feed, and so on.
So the bottom line is how does it compare to say a 10 buck an hour human?
Well, it’s 100,000 bucks, so the same as 10,000 human hours, and 40 hours a week for 50 weeks is 2,000 hours, so five years human wages.
In five years the robot is probably going to be obsolete or worn out or both.
At 20 bucks an hour you are comparing to 2.5 years of human labour, so the numbers get better.
If the parts being worked on cost 100 bucks each and robot wastage is 0.05% and human wastage is 0.5% then ruining 1000 is your purchase cost, and the human will ruin them 10 times as fast as the robot, so if you are doing 6 parts an hour that’s 6 x 40 x 50 = 12,000 per year, 0.5% wastage is 60 wasted by the human vs 6 wasted by the robot, so 5,400 bucks saved by the robot in one year.
chomp up to 60 parts an hour and you get 54,000 bucks saved by the robot over the human.
one small back injury and a couple weeks off work, at work, slipped on a floor, can cost you 100k easy.
So the answers are complex, and the robot is still the domain of the repetitive process, ideally some process that is done at least 25,000 times before you even think about re training.
25,000 on a production run is a lot if you’re talking motorcycles, nothing if you are talking brochures.
And this is usually where the whole robotic / automation discussion ends, it’s for mass production or nothing.
Really?
While the retraining and new grippers for each job arguments are valid, they can be deceiving too.
You might be able to use the same gripper for every single job, because all it carries is a spray gun, and you might be able to re-use / reload old training files, because you only need one for VW beetle front nearside wings, one for VW beetle rear offside wings, etc etc.
It doesn’t have to be a spray gun, it could be a milling spindle, or a 3D extruder, or a laser, or a plasma cutter, or a mig welder, I know (anecdotally) of one such robot that pretty much does nothing but apply glue to composite laminates, it’s a process that has to be completed perfectly in between 120 and 240 seconds, so the glue can make the strongest bond possible, the robot allowed them to boost yield from 10/15% to 70+%
But even here, none of this shit matters, what matters is the robot that costs you 100k now will cost 75k next year, and 60k the year after, and 50k the year after, meanwhile the cost of employing the human will go up.
In principle 6 axis robots are simple enough, just servos and toothed gates style drive belts in housings with bearings to allow movement, the trick and the expense comes with the repeatable precision, and now you can buy robots that do not need a cage but which can share a working space with humans.
It’s called “collaborative robotic technology” and it’s basically a fucking game changer, like in that clip, if it senses the human finger in the way, it’s a piece of piss to sense something else wrong like insertion force, and you go down to 0.01 % wastage or less.
We are not there yet, but we are oh so fucking close to the point where you could tip a lego kit into a parts bin, train the robot once, and then have it assemble the parts into a working model, time after time after time, and by oh so fucking close I mean you could buy one for US$100k.
You could buy one today for a couple of million.
At that point you have a robot that can knit, or sew, or assemble a small IC engine, or any one of a million other boring repetitive tasks, and the only thing new need humans for is to bring up new parts bins, and remove completed parts bins, and that is something amazon is doing today with its warehouse robots.
And suddenly you can eliminate the human picker and packer at the automated amazon warehouse, and for an operation like that that runs 24/7 100k is chump change compared to the churn of agency staff.
For a very small high tech business like the one I own and run, the argument for the robot in preference to the human worker is not quite there yet, but an order for 50,000 identical parts would change that.
That fact alone should concern you, because it might serve to open your eyes to the true fact that my small business already employs robots, they just aren’t six axis so they do not look like robots, so the six axis collaborative high precision shit is just one new layer on a cake that is already there, and that makes it fucking inevitable.
I simply cannot imagine *anyone* buying a 100% manually operate mill or lathe today, with the *sole* exception of a single individual buying some old commercial machine at basically scrap prices, because it is old and rigid and fairly butch and still accurate enough for the hobby shit.
So if nobody is buying manual mills or lathes, nobody is looking to apprentice or train or employ old school turners.
The job is gone.
The job went when the machines that required it went, NOT the other way around.
Word processors did NOT come in to compensate for the lack of typists to fill the typing pool.
You do not see all that has gone, the humble laser printer eradicated the print and litho works down in the basement, email systems eradicated the mail room, word processors eradicated the typing pool, accounts systems eradicated the clerks departments, barcode scanners and inventory systems eradicated the stock managers and men, and the beat goes on.
I was alive when a computer was a multi million dollar thing that weighed tons and filled small rooms and that was as good as it got, within a dozen feet of me are two PC’s one laptop, 4 smartphones a couple of arduionos and some cnc kit, and I own it all myself.
Now we are at the border of precision human collaborative six axis motion, and it is already at the point where if you can employ a human you can just about afford it, so ten years at best before if you can afford it and it can do a particular job that a human does just as well, then you can’t afford to employ the human.
Men, particularly the type of men that read this blog, get all this instinctively, every single one of you without exception has tech tools and wants to buy more, as soon as they get better and more affordable.
