Link to intringing McKinsey report on affects of future automation (employees, average)

[bobtn]

I've just read the summary, and it looks like interesting stuff. Friday, I'm going to download the full report..which is linked here. It is 176 pages long, so it will download faster for me at work.


Disruptive technologies: Advances that will transform life, business, and the global economy | McKinsey & Company

[andywire]

Companies live and die by their capital investment choices. I've seen a lot of companies hit the auction blocks with an eye bulging assortment of cutting edge, high tech equipment. I think we have a fair ways to go yet. A lot of this robotic stuff, along with automation is great if you're trying to make 1/2 million widgets. The real question is, does 1/2 million widgets at 10 cents a piece justify a million dollars in capital expenses? Doing the math, it may not. The fact is, the current cost of a lot of this technology does not warrant it's use. The real question is, at what point does innovation and practical application converge, and hopefully make a profit? Many companies have tried to find out blindly, and have gone out of business just as fast.

As for most of the work done outside of large scale production, we are a long ways away from eliminating the worker. The first 5,000 widgets will always be a test run, where troubleshoot will be required to solve the inevitable problems. We already have cutting edge technology capable of cranking out work fast. We've always had technology like this, as machines from the 40's were actually faster than most of what's available today in terms of productivity. The difference is we can meet increasingly more challenging and sophisticated geometrical and dimensional demands, which used to require more workers. So, we have a wider pool of total work that we can, for all intents and purposes, automate.

There are companies today who are making large profits with old technology. The difference is the type of work done. Small scale production yields larger per widget prices because both new and old technology requires nearly the same human set up time. We are quite a ways off from drawing a picture and making that piece with no human intervention. There's always problems in between, there's always a demand for experienced troubleshooters and machines cannot think or problem solve. They always follow a ridged set of preprogrammed commands. The best one can achieve is to set a string of commands such as "if"... followed by "than". Unfortunately, computers can only think in terms of their programmed parameters, and never outside them. Teach a computer to think, and then you might be on to something big. For now, we are limited to finding someone who can anticipate every little potential problem. They are a rarity.

Everybody's talking about 3D printing these days. Most are shocked to find out it's been around since the 80's. It wasn't practical or cost effective then, and I wouldn't be surprised if that is still true today.

[Bo]

Here's the truly scary part for anyone involved in a job they thought could never be automated.

Advanced robotics—that is, increasingly capable robots or robotic tools, with enhanced “senses,” dexterity, and intelligence—can take on tasks once thought too delicate or uneconomical to automate. These technologies can also generate significant societal benefits, including robotic surgical systems that make procedures less invasive, as well as robotic prosthetics and “exoskeletons” that restore functions of amputees and the elderly.

If surgery can be automated, no one in a "skilled" position should consider their job to be safe forever.

I've read elsewhere that the next big disruption could be in transportation, as robots could do the work all sorts of positions that involve driving.

[bobtn]

Originally Posted by Bo

Here's the truly scary part for anyone involved in a job they thought could never be automated.



If surgery can be automated, no one in a "skilled" position should consider their job to be safe forever.

I've read elsewhere that the next big disruption could be in transportation, as robots could do the work all sorts of positions that involve driving.

Correct. What it comes down to is the cost of automation is dropping like a rock at the same time the cost of a human is rising faster than their paycheck because Health Care inflation is several times gross pay inflation.

Corps buy based on ROI; ROI is simply cost/annual savings. Most corps past the mom and pop stage have a maximum quantity of years as the limit to purchasing automation. Denominator decreasing / numerator increasing decreases ROI years fast. (We've automated stuff that would not pass the ROI test just a few years ago, for that reason. Cost down, savings up).

At some point not too many years away, I'd bet we'll see automation reducing jobs faster than we did during Q4 of 2012.

You are spot on regarding driving; states are already writing driving regs in response to automated car technology updates. That means they see it as a near future occurrence.

[andywire]

Originally Posted by bobtn

Correct. What it comes down to is the cost of automation is dropping like a rock at the same time the cost of a human is rising faster than their paycheck because Health Care inflation is several times gross pay inflation.

Corps buy based on ROI; ROI is simply cost/annual savings. Most corps past the mom and pop stage have a maximum quantity of years as the limit to purchasing automation. Denominator decreasing / numerator increasing decreases ROI years fast. (We've automated stuff that would not pass the ROI test just a few years ago, for that reason. Cost down, savings up).

I could definitely see this as being practical in a large production setting. For the most part though, I think much of the job loss would be low skill occupations like shipping/receiving, packing, counting, cleaning parts before packing, etc. Dexterity is something no robot currently offers, and most automation experts have told me that's many years, likely decades away. You would have to somehow incorporate a human's level of dexterity into the parameters, a daunting task I could only imagine. Essentially, you might even have to teach it to think in some fashion.

And really, it's not just large corps that are investing big in automation. The technology works well for anyone who specializes in large scale production, including many mom and pop operations. We subcontract work out to a local swiss lathe shop that currently has 50 machines, and the owner has about 3 employees (one for each shift) working for him. He only takes in work in quantities of over 100,000 pieces, and no bigger than an 1/8th of an inch. The material is pulled through each machine off large spools, and workers are only required to change spools when needed, and occasionally check work. Pieces of that size average anywhere from 8-12 cents a piece, and the place runs 24/7.

I don't see his headcount dropping any more in the coming years no matter how much automated technology is presented. I've actually worked in his operation on an as need basis, and it's quite cutting edge. When everything runs smoothly, he makes upwards of 10K a day in actual sales, although his operation is heavily financed into the 10's of millions. One nasty recession could literally wipe his business out.

One area that I see robots already taking over is inspection. At my previous job, I assisted with the construction of a machine used to assemble plasma cutter tips. As the pieces ran through the machine, a camera would inspect every piece prior to the addition of a rubber O-ring by a robotic process. If the piece was bad, it would be knocked off the conveyer and into a collection bin. Unfortunately, this did nothing to prevent the production of bad parts, and the company refused to pay enough to hire workers capable of running highly productive shifts. Lots of high tech, expensive gadgets, yet all is for not without qualified labor.

There are plenty of ways to reduce the number of workers needed, but that also takes workers to figure out the issues. I think so many jobs have been eliminated already that many employers are having difficulty thinking of ways to further reduce headcounts. There are experts in the field who work as consultants, but they charge a steep fee and their results come with no guarantee. In many cases, they are contracted to perform their services in places where automation and robotics currently is not economical. More than a few companies have lost big money attempting to go this route.