Continuous Flow

The concept of flow production is not new. Henry Ford was the one who originated this system, through constant innovation, in the early days of the industrialization of mass vehicle production, at the Highland Park plant - in Detroit - Michigan - United States, in 1913, with the introduction of the moving assembly line.

This drastically reduced the “lead time” and the manufacturing cost. To give an idea of the impact, the cost reduction exceeded 50% and the time to assemble a vehicle was cut from 12 hours to 90 minutes.

Ford's flow system revolutionized industrial society to such an extent that, even today, its influence can be seen in practically every manufacturing company with serial production.

In 1950, when Toyota needed to build vehicles faster, in small quantities and at low cost, it drew inspiration from many things in the Ford system, especially the moving assembly line, to structure TPS. The idea was to extend the concept of the moving assembly line to other production processes in the system, giving rise to production in Continuous Flow cells, which we can say is the successor to the moving assembly line developed by Ford.

Although the fastest way to transform RM (raw material) into FG (finished goods) is continuous flow, historically companies have adopted production with a process-based layout organized by machine type to carry out mass production. It seems to me that the human mind intuitively likes to produce in batches.

The Continuous Flow process aims to eliminate the production stops and restarts that are typical of a conventional system. This shortens the “lead time” by significantly reducing non-processing time, eliminates WIP (work in process) and also helps maintain quality by immediately detecting nonconformities.

To process effectively in Continuous Flow, the premise is to have stable processes (see the article on Stability already published in this column), and it is necessary to produce one item, only one item, and pass it immediately to the next process.

Batch production is not allowed, even though one may have the feeling that batch production increases efficiency in the transport process. On the contrary, all processes must direct only one piece at a time, corresponding to the single unit arriving at the end of the assembly line.

Some production processes, however, require extensive and time-consuming tool changes (set-up) every time a different part needs to be produced, thus making one-piece-at-a-time production impossible. In situations like these, typically in stamping, plastic injection molding and various machining operations, the batch size must be as small as possible, aiming at one-piece-at-a-time production, even if it is very difficult. Reducing batch size necessarily means reducing set-up time.

Toyota has worked very hard to reduce set-up time in all processes. For example, in 1970 the set-up time on a stamping line with 1,000-ton presses took more than 4 hours; after three years of constant improvement it was reduced to less than ten minutes. Today, the set-up time on the set of presses that have 40- to 50-ton dies to be changed is under 5 minutes.

Let us now look at the steps to create Continuous Flow:

1. Define the product families through a detailed analysis of the operations for each item.

2. Whenever possible, align all product families in the sequence of the processes.

3. Observe critically, with a focus on Continuous Flow, and do kaizen on:

- discrepancy between takt time and actual production: first analyze the PAB (Production Analysis Board) and check whether there is a discrepancy between planned and actual production. Any divergence, that is, above or below, means there is a problem and requires immediate kaizen. Takt time must be followed rigorously;

- small stops between operations: whenever there is any stop, however small, between operations, there is a problem. Search for the root cause and do kaizen;

- motion waste: if the worker's movements do not repeat every takt time, it means there is batch production and not one-piece flow. Waiting time, in most cases, is caused by the segmentation of operations. Motion waste indicates that the positioning - physical position - of the parts has a problem. One way to check whether there is a problem with the worker is to compare the current layout with the one-piece flow layout.

In addition to the steps above, there are other factors that must be taken into consideration:

1. Synchronized Production: by stabilizing and leveling the production capacity of each process, WIP (work in process) is eliminated.
2. Multi-process handling: one worker operates several machines in the sequence of the production flow within takt time; avoid the waste of overproduction.
3. Multi-skilled worker: in order to carry out one-piece flow production and multi-process handling, the worker must be able to perform several different tasks and operate several machines and pieces of equipment guided by takt time.
4. Arrangement of machines and equipment in process sequence: U-shaped layout.
5. Machines and equipment that enable Continuous Flow: acquire machines and equipment endowed with “intelligence”.

Before closing, I would like to recall and emphasize that Continuous Flow is one of the essential elements of the JIT (Just In Time) pillar, together with Takt Time and the Pull System.

The goal is to produce and move one item (part) at a time through a series of processing steps, continuously, so that at each step only what is required by the next step is done.

In the next article we will discuss the Pull System.

Transform your plant with Lean manufacturing consulting. Improve processes, reduce waste and achieve greater efficiency at every stage of production. Achieve real and sustainable results. Visit and find out how!