Automation in Chinese factories: not a solution!

Process ImprovementEvery time I read an article about rising Chinese labor costs, it seems there are only two solutions: moving to inland provinces or automating production. I do not really buy into these proposals, and I want to focus on the myth of automation.

Automating a process allows a reduction of headcount (lots of operators are replaced by a few technicians who set up the machines) and an acceleration of cycles. But there several drawbacks.

First, it is a large upfront investment. Does it make sense to make 200,000RMB molds rather than 20,000 RMB ones, when every customer insists on making his own designs and on renewing them every year? No, except for very large and predictable volumes.

Second, automating a poor production system does not make it better. It improves some efficiency measures (e.g. direct labor cost), but it does not address the most important opportunity for savings–the optimization of the whole system.

This second argument is backed up in a VERY interesting article on Evolving excellence that describes the most profitable Toyota factory:

There was not a computer visible on the floor. Anywhere. A large electronic sign that showed the status of each line, but everything else was manual. Individual parts bins were manual kanbans (but with bar codes so the paper cards themselves didn’t transfer). When empty they were replenished by guys driving trains of carts up and down the aisles between the lines. As a bin of parts was emptied, another was put in its place. For thousands of bins, parts ranging in size from washers all the way up to engines and doors, across acres of factory floor.

Toyota automates only what is dangerous (welding) or too heavy for humans. Everything else is done by humans… because humans can create ideas for improvement. And that ongoing improvement is one of the truly impressive aspects of this operation.

From my observations, 99%+ of Chinese factories can easily shave 10% of their costs, simply by applying some good manufacturing common sense / best practices. No need to invest millions in new toys.

What do you think?


  1. says

    Yes, they Chinese would be much better off using ‘Toyota common sense’. I’ve seen HK ‘managed’ factories get >10% quality/quantity just by using a better mix of HK and PRC culture in the process. Take advantage of the people skills, and keep them around long enough to grow the system.

    I’ve found they listen on showy how-it-sell features, and blank wall into a numbers game when talk roams into production/quality assurance.

  2. says

    Very good point. I also fully agree. The best factories are not the ones that have the best hardware but the ones that put in place good manufacturing practices and keep using them in the long run, then continue improving every time.

    And this is certainly reducing cost. Chinese suppliers typically do not know the cost they pay for rework. I participated to a China Supply Chain Council seminar on quality a few months ago and there was a very interesting speaker. He explained how he is now able to sell quality improvement training services to Chinese suppliers, not because it brings higher quality (which these owners did not see as a must) but because he could show it would cut manufacturing cost.

  3. Renaud Anjoran says

    Hi Etienne,
    Yep, the cost of non-quality is often quite significant for Chinese factories. Re-working defective goods alone is easily 5% of their costs.
    There are also other costs… What about exceedingly long production cycles, degradation/loss of work-in-process goods that are stocked on the side, unexpected delays that cause penalties/extra freight costs, and above all a poor mapping and execution of processes?

  4. says

    Renaud – couldn’t agree more with your post. Good points. Automation can help in the long-term, but most factories could save 10% off the top with some simple restructuring of the production areas. Your posts are great keep it up!

  5. Twofish says

    Just to bring in some historical background. There is a reason why the industrial revolution happened in Britain and not China. China was and is labor-rich but energy-poor, whereas Britain was and is labor-poor but energy-rich, so it turns out that because of the high population in China, there was not (and is not) the economic incentive to use capital to save labor.

    The British in the early 19th century wanted to open up Chinese markets so that they could keep the looms of Manchester spinning. They never had a chance, because most British textiles are made in China. Textiles happen to be something for which high capital automation works rather badly.

    One problem with automation is that you have cheaper per unit costs, but you pay for it because you don’t have nearly the flexibility that you have with labor intensive systems. For example, think of cars. It would be nice if you could go online, order a red car with exactly the right engine and exactly the options you want, but you can’t because heavy industry is just not that flexible.

    Also, the really big change is going to happen around 2020 although you can see signs of it now, as the one-child generation starts going into the workforce and the baby boom generation starts retiring.

  6. Renaud Anjoran says

    Twofish: very true. Actually some car factories can deal with thousands of combinations and make to orders, but they are not necessarily the most automated.