Robots vs Manual Labour: Cost & Efficiency Comparison

Mark Tickner said:

So the question is, how did society change back then and does it have implications for any further changes. But I agree, most major manufacturing I expect to already be utilising robotic labour.

On Radio 4 this morning, in a very good programme about trade unions, the point was made that there's lots of fuss being made about AI making jobs redundant because now it's white collar jobs that are under threat...

Mark Tickner said:

That question is arguably a few decades too late. 

To get unusually serious for me, it certainly should have been asked a few decades ago, but we really need to hope it's not too late. Let's not get into specific politics, but the lack of employment opportunities is a perfectly understandable reason why there's been societal concerns across many or most Westernised counties, which of course have been played on.

It's a really tough question in a free market economy (which is why no-one's found a solution yet), it's not the responsibility of private or even public companies to provide employment. Private companies responsibility is to make a profit, public companies to provide a service. 

There is one area here which it feels to me should be solveable, and does come within the remit of this forum, why is it so difficult to get a job as an electrician while at the same time it is so difficult as a householder to locate a reliable electrician? (Ditto anyone else in the building trades.) My suspicion has long been that it's been the lack of companies able to provide apprenticeships because most are too small to be able to afford the time and cost, is that a reason and are there others?

There is a scale issue - in electronics, if you are 'going large' you invest a few million and get a custom chipset made, - the Apple/Intel sort of approach, or even Panasonic/ Toshiba, following in a path blazed in the 1970s by Mulllard/ Phillips. 

Modern chip making is both highly automated, when you want a thousand wafers of chips per year, and highly manual, if you scarcely need one wafer.

If you are not in the multi million financial depth of the custom chip, you can only stick parts made by others on a PCB. In an Research setting where you only want one or two, that may be a totally manual process, but even then, the pick and place machine will be used for small surface mount parts, and if you want more than ten of something it is automated.

What really cannot yet be automated is the final fix, especially for things that get installed in houses and gardens, which are also all different. 

for the same reason the bricks were laid by hand, the light switches have to fitted by hand, pipes for the loo fitted by hand etc but of course power tools and jigs make this as fast as possible, and generally the parts end up far cheaper than the labour for a lot of it.

In the end the jobs that remain are the ones where every case is different - and that means  hand on things like dentistry, nursing and perhaps hair dressing, and fitting things into odd locations, construction, plumbing and house wiring are good examples of that.

The step and  repeat jobs like putting lids on toothpaste have been automated all my life, and I am nearly retired, and there has been a gentle creep towards more complex things, and many processes now need cameras on computer to do final inspection or to line things up.

Its not a simple yes no question.

And the ideal of us all having more free time to pursue fine arts music and so on seems as far away as ever, we now just have far more stuff, and no time to play with it !

Mike.

You are right, due to the fact that multinational companies sometimes transfer that manufacturing towards country where labor is cheaper than robots!  

Hi Alan, it is very useful the "Guide to implementing Industrial Robots"!.... Of course I will read it"... thanks indeed!

Andy Millar said:

certainly should have been asked a few decades ago

Maybe we should review the idea of de-skilling pin manufacturing. There was a book about it, apparently. Some guy called Smith, a long while back. Mr Ludd was unhappy.

hi Antonio Leone in collaboration with IET Coventry there is a talk on the 22-Apr at 1900 about the guide mentioned above. It's intended for engineers, supervisors and managers in manufacture.

link and registration are below. all welcome.

• IET Main Calendar Listing: https://events.theiet.org/events/industrial-robots-benefits-barriers-and-the-business-case-for-the-implementation/
• Event Registration Page: https://localevents.theiet.org/register.php?event=4b5a0e

#robotics

Hi Alan, thanks indeed.... just registered!

Bang on. It's very task dependant. Some dexterous, variable and multi-stage tasks are far less appealing to automate, as they have a high cost and may be even slower than a person to carry out (particularly is inspection for success is required, something a human worker can do by-the-by).
I believe, if manpower is a serious issue in product throughput (and in the UK it typically is!), then automation is the a great way to multiply how much work you can do with limited staff. You look for the biggest wins first, and sell workers on it by showing how it can be a tool for them, rather than a replacement. After that, tough as it is to swallow, it might be worth looking at evaluating a product design to see if it could be made easier to automate, particularly at bottlenecks in the line. And that's all assuming it's not much easier to just leave it as is and got abroad for cheap and plentiful manual labour.