Lean without Jidoka is just Fordism
Why do you adopt lean practices? Is it to try to create ever-better products? Or is it to improve flow and increase productivity?
If you look more closely at most lean programs, you will find behind them the thinking of Goldratt's The Goal: if we improve flow (by removing bottlenecks), we increase productive capacity and reduce the use of resources, all in order to make more money (which, according to him, is THE goal of any business).
Sakichi Toyoda was observing the human hardships of cotton production when he sought to reduce both defects (problems in use by operators) and excessive effort (difficulties in doing the work). It was in this context that he mechanized the loom, inventing the automatic shuttle change, and created the famous concept of jidoka, with thin metal strips that tested the tension of the cotton thread and stopped the machine if the thread broke. He was solving human problems with automatic devices. His goal was to create ever-better looms, just as, later, Kiichiro Toyoda's goal would be to create ever-better automobiles.
What does this mean, concretely?
Our society is built on the creation of mechanical devices that transform our way of life by replacing human labor in activities ranging from transportation (trains), through the textile industry (sewing machines and also spinning equipment), to agriculture (tractors). The purpose of these products is to make you autonomous – that is, to no longer need to do the work directly, nor ask someone else to do it for you. Think of espresso machines, now ubiquitous in offices around the world: before, you had to go to a coffee shop to have a coffee; now, with these machines and capsules, you can make one on your own, without anyone's help. The same goes for washing machines, dishwashers, word processors and so on.
Even so, for every device, there are two dimensions driven by humans:
- Product: the mechanical tool you can use autonomously to get what you want.
- Service: the person you want to talk to when you have problems with the machine.
Autonomy, with any device, requires a level of skill proportional to the difficulty of what you intend to do with the machine. Most products, after the initial – and almost always unpleasant – phase of “getting started” (that stage where most of us muddle through, grumbling, reading the manual), can be used with minimal skills in simple cases, but require much more knowledge in more complex usage situations. At that point, the customer has two options: learn to do the work on their own – which may not be their preferred way of investing their time – or find someone to do it for them and show them how it's done.
Managers obsessed with operational efficiency believe that the goal of automation is to eliminate human resource costs. They dream of fully automated systems that customers can use without ever needing to talk to anyone. Then they don't understand why customers scorn their offering, look for alternatives and abandon the brand as soon as they get the chance. Customers are not looking for a specific product; they want a neat solution to their problems. They don't just want any coffee – they want one that is simple to make, tasty, comes out of a stylish machine, is easy to enjoy with friends and comes with quick access to support when something goes wrong.
The great insight of jidoka is that mechanical devices are just that – mechanical. Because they lack human judgment, they inevitably:
- Produce defects – how could they know what level of quality a human being expects (something that is also contextual)?
- Break down – how could they monitor and replace their own parts?
- Never learn to create the next generation of better machines.
Every piece of equipment produced has a service attached to it. This requires someone to help the user with:
- Using the machine correctly to obtain the desired results, according to the level of complexity of the use.
- Fixing the machine when it fails.
- Improving the machine, creating a model that is smarter, easier to use and better looking.
Jidoka is about developing the human part of a product. We don't need people just to watch machines; we need them to identify customers' problems and solve them courteously and efficiently. Having self-checkout machines in a department store is a smart idea, as long as it is understood that someone will still need to be present and trained in customer service (first impression, availability, tone of voice, problem solving, etiquette, etc.) to create a smooth and pleasant experience for customers.
If this person is trained in jidoka, they will also be able to recognize all the problems of the automated system and work with the engineers to improve the next version. This is the entry point for value analysis (correcting flaws in machines already in production) and for value engineering (perfecting the next generation of the product).
From Sakichi's point of view, the goal of a company would not be to make money, but to benefit society by helping people do things autonomously, through products that are continuously improved. Seeking to improve flow and productivity by reducing lead times is an excellent way to reveal production scheduling and logistics problems, but it contributes little to creating ever-better products. For that, we need jidoka: the ability to detect defects or wear; to understand where the process deviates from standard work; to stop the line rather than let the problem worsen; to have management react quickly to analyze and correct the failure; and to incorporate this correction into the workflow and, later, into the next generation of machines.
Jidoka is more than a way of running production lines smoothly and reducing the cost of carrying defects through the flow. It is a fundamental way of teaching people about products and equipment, so that they better understand the relationship between automation, human contact and problem solving. Jidoka is about understanding the appropriate role for humans (never use a person to do a machine's work), for machines (give them the capacity for self-assessment) and for the interaction between the two (teach humans to diagnose problems and design better machines).
In the age of Artificial Intelligence, as we automate intellectual tasks that we once considered exclusive to human beings, we will need jidoka skills — and very much so — to redefine where work should be human and where it should belong to the machine. Even so, even if you use ChatGPT to write all your code (or your essays), you will still face problems related to how customers use the system generated (or how the teacher grades the text), how AI generates nonsense or loses context, and how to sharpen judgment in the design of what you do.
In this era of division of labor, it is easy to reduce lean to “that production thing” and apply its techniques only to improve flow, increase productivity and – perhaps – save some resources in the process. However, the true goal of lean thinking is to create ever-better products, at affordable prices and with lower production costs. Certainly, just-in-time is part of this, but its focus is mainly on planning, organizing and moving items – not on their design. Jidoka, on the other hand, both in the ability to recognize problems and in the separation between human work and machine work, is the essential pillar for learning about the interface between product and service and, organically, developing more complete and improved offerings, with better products and services that increase customer satisfaction.
Without jidoka, lean becomes merely an extended form of Fordism – occasionally useful, but far from transformative. Ask yourself once again: “Why do I think lean?” How do you intend to help the company create ever-better products? The answer lies as much in service as in production and design – machines will never understand the ambiguous feelings and complex dilemmas that humans naturally face every day. Only humans can understand what humans are really looking for, what problems they experience and what kind of interesting solutions they want.
