The Future of Manufacturing

One result of my longstanding interest in manufacturing is that I’ve noticed certain trends. The most notable trends come from the impact of machines, both computers and robots. Global manufacturing makes much more use of robots than it once did. We can prototype objects whose design once made them infeasible. Automation continues to capture new domains, previously reserved only for human hands. I predict that these trends will continue, and they will deeply change manufacturing. In addition, I think they will have impacts on innovation, entrepreneurship, globalization, politics, and on all of the objects that make up our everyday world.

The trends I write of are a cause of worry for some people. I don’t see them as threats, though I’m not in their obvious ‘warpath,’ either, at least for now. But, like them or not, the machines are coming. Imagining what could happen may be worthwhile, for all of us.

Computers

It goes without saying that computers have changed things. Computer Aided Design (CAD) software has made a massive difference in specifying products for manufacture. Computers that interface with manufacturing lines and embedded in individual machines have increased efficiency and precision.

3D printing (or additive manufacturing) provides a new outlet for exploring different designs. For a while, 3D printing was mostly limited to prototypes. This seems to be changing rapidly. GE uses additive manufacturing to produce fuel nozzles for airplane engines, and Adidas has recently built a facility that uses additive manufacturing to build limited runs (around 500) of certain sneaker models.

While 3D printing hasn’t brought the revolution that some anticipated, it has pushed manufacturing in a new direction. Economic theory predicts that technologies will be put to their most valuable uses first. In this case, it seems 3D printing is good for making low-volume, high-value parts. As the technology improves, more things will be 3D printable.

I think the major long-term impact of 3D printing will be to allow for more distributed manufacturing. Some products that are currently made in one spot and shipped to users will simply be made by users. GE’s fuel nozzles are an example of this kind of on-site production.

Distributed manufacturing is important because it complements the distributed nature of human capital in design. The internet and personal computers have allowed anyone, anywhere, to design objects for manufacture. Recently, I paid a designer on Upwork to create 3D models for an object I’d like to manufacture. The designer lives in Ukraine, but it wouldn’t have been much different if he lived across the street from me. We never really needed to interact beyond text, pictures, and sharing files, and money, of course.

If everything were sold strictly via brick-and-mortar retail outlets, none of these trends would matter. All of the distributed design and manufacturing capacity would be pointless if big institutions could crowd out little guys at the point-of-sale. But platforms like Kickstarter, Indiegogo, Amazon, eBay, Stripe, and Shopify have cracked retail open for the little guys too.

In short, for some products, you can come up with an idea, hire a designer on Upwork to produce models, send it to a 3D printer to get a prototype, go to Kickstarter to get initial traction, and then set up a standalone site or take it to Amazon. Granted, for electronics, this still doesn’t work. You’ll still need to go to China for most hardware. But for more and more things, this is a viable route.

I see this trend continuing. If it does, it may have some very interesting implications. First, think about what I would have done if I didn’t have my options: probably nothing. I don’t have the time to divert from my main focus (research in quantitative marketing) to figure out how to get it manufactured. So as distributed manufacturing spreads, more people make stuff. This is analogous to the way more people made videos once camcorders fell in cost enough that most families could have one. Additionally, things will become feasible that weren’t before. GE again: it would have been impossible to make the fuel nozzle in any way except for additive manufacturing. New production methods open new design avenues.

Now a second implication of distributed manufacturing: some traditional processes will be replaced. My Mom recently went to the dentist and had to get a crown. Instead of having a mold made and sending away for someone to make a crown, the dentist sent a digital file to a machine in the other room and came back with the crown in minutes.

In a nutshell:

1. New manufacturing tech will allow new people to make things, and new things to be made
2. New manufacturing tech will replace some old manufacturing tech

The first effect might be called an expansion effect, and the second a replacement effect. I don’t know which one will be stronger, but it’s a really interesting question. It’s also probably possible to find out with data we have now. Maybe someone (probably an economist) has studied it already. (If you know of a paper, send it my way!).

Robots

Robots scare people. I won’t dwell on why that’s true, but I do hope to paint a picture of how robots can be great. I’m partly imagining here, partly predicting, and partly hoping.

Robots typically do have impacts on manufacturing, but they are not always the same. Three recent papers find somewhat different impacts of robots in manufacturing. In the U.S., for example, it seems robots replace jobs to a large degree, while in Germany, this is not so. Usually the hardest hit are low and medium skilled workers. Productivity increases always, however, and highly skilled workers either become better off or at least not worse off.

These empirical findings may be suggestive of what will happen as we move forward. We won’t have many low-skilled or medium-skilled workers in manufacturing. We will instead have more productive, highly-skilled workers.

If this is true, it will have a big impact on the distribution of big manufacturing the world. There is a worry about the U.S. being weak in manufacturing. One quote that captures a common idea is from Tim Cook:

China put an enormous focus on manufacturing. In what we would call, you and I would call vocational kind of skills. The U.S., over time, began to stop having as many vocational kind of skills. I mean, you can take every tool and die maker in the United States and probably put them in a room that we’re currently sitting in. In China, you would have to have multiple football fields.

If you romanticize manufacturing, like I do, and worry about losing the potential to do it competitively in the U.S., this quote is a slap in the face. But think about robots. Tool and die makers are big targets for automation, with current technologies. Over time, China’s human capital advantage will probably be rendered moot.

What will be important is who has the highly-skilled workers. That’s not particularly more comforting if you’re rooting for U.S. manufacturing, but it’s not the insurmountable situation faced in the domain of tool and die makers. The U.S. still manages to attract extremely bright people for highly skilled work, via the university system. In addition, since the U.S. has extremely expensive labor, it will likely embrace automation faster than countries with less expensive labor. Bloomberg, referring to a machine that automated a tricky step in shoe manufacturing, and used by Nike, noted that:

This could be a step forward in Nike’s attempt to change the economics of shoemaking so it can relocate manufacturing closer to the big consumer markets in the U.S. and Europe.

A second example, this one from Quartz, and referring to a new machine that can (finally) sew t-shirts:

Normally, manufacturing in the US would be much more expensive than producing in China because of the higher labor costs. But Tang Xinhong, chairman of Tianyuan Garments, told World Textile Information Network that, in a completely automated production line, the cost of human labor works out to about $0.33 per shirt. For context, to produce something like a denim shirt in Bangladesh, you might pay about $0.22 in labor costs, according to an estimate from the Institute for Global Labour and Human Rights. That same labor would be $7.47 in the US, putting the labor cost for Tianyuan Garments’ American-made shirt about on par with one of the cheapest labor markets in the world.

Where does this put the conversation about manufacturing in the U.S. vs. out? Well, it’s true that China, for one, believes automation is the future, and the decision makers there are pushing for more of it all the time. In fact, one of the first places Nike build the machine referred to above is in China. But somehow, the question of whether a U.S. robot makes something vs. a Chinese one doesn’t seem as significant as whether a U.S. citizen or a Chinese one does.

With robots, we become unmoored from the building of things, and we spend our time on the higher-order tasks, like design. This, to me, is actually quite hopeful. The higher-order tasks tend to be the more fun and fulfilling ones. In any case, this shift to a new kind of skill in manufacturing will probably lead to a shift in where manufacturing mostly takes place.

One point unrelated to manufacturing is worth raising here: we should probably setup investment vehicles that the average person can use to hedge against automation risk. If you are a tool and die maker right now, it’d be nice to own an asset that will increase in value if new technology arises to replace you. I have neither the financial expertise nor the time to construct and market such an asset, but if someone reading this does, it sounds like an idea worth vetting.

Robots, then, will probably have these impacts:

1. Manufacturing centers in the future will be the places where human capital and labor costs best support automation. This may continue to be China, but the U.S. isn’t as pathetically far behind in this domain.
2. The question of which humans know how to make things will be replaced by questions about which humans own and operate the robots that make things.
3. Combined with distributed manufacturing, above, robots will probably allow the ‘little guys’ to not only prototype and sell in small quantities, but to scale up, as well. Especially if clever people create more widely available access points for robotic manufacturing.

Humans In a New World

I could be totally wrong about all of this. But if I’m not, there are a few things that might be worth doing, depending on how you want to participate in manufacturing.

1. If you want to work in manufacturing, build human capital in the high-skill aspects, like design and engineering.
2. If you want to a physical, non-electronic, product you can use the chain of services I detailed above. There is something thrilling about taking an idea from concept to finished product, and even more thrilling if you can sell some.
3. If you want to financially benefit from trends in manufacturing, invest in robots, design software, and new tech that expands distributed manufacturing.
4. If you want to play a role in making more things in the U.S., maybe you can have some hope now. Though, any new U.S. manufacturing probably won’t resemble the kind we used to have.