I am getting ready to teach another iteration of the Kellogg core MBA class. One of the first points we make in the core is that there is no one best way. That is, what is the best process for executing some bit of work depends on the needs of the business. Whether it is best to rely on lots of capital or lots of labor depends on a myriad of factors. Case in point, Japanese bullet trains.
Few things epitomize modernity or high-end engineering like trains that move along at 200 miles per hour. However, that does not necessarily mean that the trains — or at least their various components — are made using high-tech processes. In particular, it turns out that the noses of the trains are made by a team of eight craftsmen who bang sheets aluminum with hammers — yes, hammers — and then weld the pieces together. (See A hammer — yes, that low-tech tool — helps mold noses of Japan’s bullet trains, Mar 28, Washington Post.) The craftsmen in question work for Yamashita Kogyosho Co, a subcontractor for Hitachi based in Kudamatsu. The company has built noses for six generations of bullet trains, including a prototype of East Japan Railway’s E6 pictured above. Here is what the actual process looks like:
The nose of a bullet train is not particularly well-suited to the expensive and highly specialized mass-production machinery that molds and cuts metal to make hundreds of thousands of cars, trucks and toasters. The number of high-speed locomotives built for each bullet-train series in Japan is quite limited, from 40 to 120. In Yamashita’s small factory, metal workers pound together a new nose every week or so. There are other ways to make one, but Yamashita’s method is flexible, reliable and relatively cheap. When engineers demand sudden design changes, the company does not have to rebuild elaborate machines. Workers simply pound out new shapes. … “The most cost-efficient way of transferring computer-assisted 3-D design to metal is with a hammer,” said Tatsuto Yamashita.
So Yamashita Kogyosho runs a job shop because they face a market that demands customization in low volume. A job shop is the classic solution. They also face one of the classic challenges of running a job shop, coming up with enough skilled labor. It apparently takes ten years to master banging out bullet train noses and their current workforce is aging. They need to bring in fresh blood if they are going to be around for future bullet trains.
In a lot of ways this makes sense. Pretty much any other way of shaping the metal would require dies and fixtures that would be expensive to develop. A manual approach minimizes the fixed costs and the risk if total production is small. I, however, can’t help but wonder how consistent they are. For example, can they can actually produce interchangeable parts? If you take the nose off, say, one 5th generation train, will it fit on another one? I’ve got to think that the railway wants as much flexibility as possible and would prefer 40 nearly identical trains as opposed to 40 unique pieces of art.