Toyota Production System





Toyota Production System

Toyota is one of the most profitable large auto manufacturers. It makes excellent products, grows fast, has high margins, and makes lots of money. Toyota's goal of going fast is not achieved by straining. Toyota eliminates wasted effort at every step of the process of producing cars. If you eliminate enough waste, soon you go faster than the people who are just trying to go fast.

This chapter focuses on the part of the Toyota Production System (TPS) that actually manufactures cars. Mary and Tom Poppendieck write extensively about the importance of the product development part of TPS to the whole task of delivering value through software.

Their alternative social structure of work is critical to the success of TPS. Every worker is responsible for the whole production line. If anyone spots a defect he pulls a cord that stops the whole line. All the resources of the line are applied to finding the root cause of the problem and fixing it. At first, American workers can't believe this. Chet Hendrickson told me the story of his brother-in-law who worked at a Toyota plant in Kentucky. He saw a defective door go by. His buddy said, "Pull the cord." Nah, he didn't want to get in trouble. Another defective door. Another. Finally, he pulled the cord. He was praised for telling the truth and pointing out flaws. Unlike mass-production lines where someone "down the line" is responsible for quality, in TPS the goal is to make the quality of the line good enough that there is no need for downstream quality assurance. This implies that everyone is responsible for quality.

Individual workers have a lot of say in how work is performed and improved in TPS. Waste is eliminated through kaizen (continuous improvement) events. Workers identify a source of waste, either quality problems or inefficiency. Then, they take the lead in analyzing the problem, performing experiments, and standardizing the results.

TPS eliminates the rigid social stratification found in Taylorist factories. Industrial engineers begin their careers working on the line and always spend a considerable amount of time in the factory. Ordinary workers perform routine maintenance instead of an elite caste of technicians. There is no separate quality organization. The whole organization is a quality organization.

Workers are accountable for the quality of their work because the parts they create are immediately put to use by the next step in the line. This direct coupling of functions was completely counterintuitive to me when I first read about it. I thought that for a mass-production factory to run smoothly there must be a large inventory of parts between any two steps of the process. That way, if the upstream machine stops working, the downstream machine can keep chugging away on the buffer of parts.

TPS turns this thinking on its head. While individual machines may work more smoothly with lots of "work-in-progress" inventory, the factory looked at as a whole doesn't work as well. If you use a part immediately you get the value of the part itself as well as information about whether the upstream machine is working correctly. This view, that parts aren't just parts but also information about their making, leads to pressure to keep all the machines in a line working smoothly and also provides the information necessary to keep the machines working smoothly.

Taiichi Ohno, the spiritual leader of TPS, says the greatest waste is the waste of overproduction. If you make something and can't sell it, the effort that went into making it is lost. If you make something internally in the line and don't use it immediately its information value evaporates. There are also storage costs: you have to haul it to a warehouse; track it while it is there; polish the rust off it when you take it back out again; and risk that you'll never use it at all, in which case you have to pay to haul it away.

Software development is full of the waste of overproduction: fat requirements documents that rapidly grow obsolete; elaborate architectures that are never used; code that goes months without being integrated, tested, and executed in a production environment; and documentation no one reads until it is irrelevant or misleading. While all of these activities are important to software development, we need to use their output immediately in order to get the feedback we need to eliminate waste.

Requirements gathering, for instance, will not improve by having ever more elaborate requirements-gathering processes but by shortening the path between the production of requirements detail and the deployment of the software specified. Using requirements detail immediately implies that requirements gathering isn't a phase that produces a static document; but an activity producing detail, just before it is needed, throughout development.

Many other aspects of TPS also have strong parallels to software development; useful concepts such as cross-training workers, organizing the factory into cells, and writing gain-sharing contracts between customers and suppliers. If you are interested, I recommend that you start with Ohno's Toyota Production System.


     Python   SQL   Java   php   Perl 
     game development   web development   internet   *nix   graphics   hardware 
     telecommunications   C++ 
     Flash   Active Directory   Windows