Traditional Mass Production Thinking
What is the ideal way to organize your equipment and processes? In traditional mass production thinking (the way most companies are organized), the answer seems obvious: group similar machines and similarly skilled people together. So mass production thinking sets up departments of mechanical engineering, electrical engineering, accounting, purchasing, and manufacturing as well as departments for stamping, welding, wire soldering, assembly, and the like. The following were the perceived benefits of grouping similarly skilled people and equipment together:
- Economies of scale. First and foremost, mass production thinking was about squeezing the most production possible at the lowest cost per unit out of every piece of equipment or every worker in a manual operation. Having one huge stamping press to meet the needs of all the factory’s products would lead to the smallest capital cost per piece. And then you wanted to run that press flat-out 100% of the time to get the greatest asset utilization. Similarly, by organizing people into departments, you can focus on best practices in each professional specialty and squeeze the highest productivity (or innovation) possible out of each person.
- Apparent flexibility in scheduling. When you put all the welders together in one department, it’s easier for the welding department manager to schedule available machines and welders to any job that comes up. If you create one-piece flow cells, you take those welding machines and welders and dedicate them to a one-piece cell, so they’re no longer free to do other work that might come up.
In mass production thinking, once you have decided to group all the similar types of people and processes together by department, the next question is how often should you move material or information between departments? Since you have organized your people and equipment by specialty, you must create another specialty, the material handling department or the planning department, to move material. That department is also measured by efficiency. The most efficient way to utilize a person moving material is to get that person to move the most material possible each trip. From the viewpoint of the material handling department, the optimal time to move material from one department to the next is when there is a large batch. The goal is to move the material once a day or, even better yet, once a week.
The best way to schedule an operation that is organized into separate pro-cesses is to send individual schedules to each individual department. For example, if schedules are developed weekly, then each department head can decide what to make each day in order to optimize equipment and utilize people for that week. A weekly schedule also provides flexibility for people missing work. You just make less that day and make it up with more production another day in the week. As long as by Friday you meet the production target, everything is OK.
Lean thinking looks at this way of organizing production and sees a company producing a lot of work-in-process (WIP) inventory. The fastest equipment, such as stamping, will build up the most WIP. Material sitting in inventory is caused by the most fundamental waste, overproduction. The mass production system guarantees overproduction in large batches, which in turn guarantees inventory sitting idle and taking up valuable plant space and, more importantly, hiding problems.
Another problem with organizing similar professional specialties and similar manufacturing equipment together into departments is that a product being made for a customer does not live in just one department. It must move across departments to become what the customer wants. Engineering, purchasing, and accounting are all located in different departments. Yet many value streams cross through these departments, causing a delay each time a process enters a new department. In a one-piece flow, you physically line up the processes in the sequence that will produce the customer’s order in the shortest time.
Illustrates a simplified view of a computer maker that is organized into three departments. One department makes computer bases, the second makes monitors and attaches them, and the third tests equipment. (Of course, in the real world there would be many departments and companies in a supply chain making a complete computer.) In this model, the material handling department decided it wants to move a batch size of 10 units at a time. Each department takes one minute per unit to do its work, so it takes 10 minutes for a batch of computers to move through each department. Even without considering material handling time to move between departments, it would therefore take 30 minutes to make and test the first batch of 10 to be shipped to the customer. And it would take 21 minutes to get out the first computer ready to ship, even though only three minutes of value-added work are needed to make that computer.
The system that Ohno set up does not assume that the ideal batch size is what is most efficient for each individual process or for the material handling department. In lean thinking, the ideal batch size is always the same—one. That is because Ohno was not trying to optimize the utilization of people and equipment in each department. When the Toyota factory was first organized, it was operating this way—like Ford’s factories. But this didn’t work, because Toyota could not compete with Ford’s volume and economies of scale. So Ohno decided to optimize the flow of material so it would move more quickly through the factory. This meant reducing batch size. And the fastest way to achieve this was to blow up departments and “process islands” and create work cells that were grouped by product, rather than by process.
Illustrates a view of the same computer-making process above, organized into a one-piece flow work cell. If Ohno were to manage this process, he would take the equipment needed to make one base from the base department, the equipment for making a monitor from the monitor department, and a test stand from the test department and then put these three processes next to each other. That is, he would have created a cell to achieve one-piece flow. Then he would have made clear that operators were not allowed to build up inventory between the three operations. For example, the computer base maker would not make the next base until the monitor maker finished building the monitor and mounting it on the last base. In other words, nobody would build more than what is needed immediately. The result is the operators in the cell take 12 minutes to make 10 computers, while the batch flow process takes 30 minutes for 10 computers. And it takes the lean process just three minutes instead of 21 minutes to make the first computer ready to ship. In fact, the three minutes is pure value-added time. What flow has done is to eliminate overproduction and inventory.