Batch vs continuous flow processing

If you are at a cocktail party, and someone asks you to compare and contrast the differences between batch and continuous flowing processes, calmly lower your martini, grab the nearest table napkin, and proceed to draw the following picture:

batch processing

Suppose we were in the toy car manufacturing business and there were three stages to building a car: the body, the wheels, and the paint. You decide the most optimal way to build the cars would be to batch them up into groups of ten, and send them from one stage to the next in one big batch.

If we assume it takes 1 min per car at each stage, then we would expect the following results:

Time to build first car: 21 minutes
Time to build first batch of ten: 30 minutes

Now, suppose we were to draw this picture on another napkin:

flow processing

Only this time instead of manufacturing our cars in groups of ten, we build them one at a time in a one-piece continuous flow process. Were we to do this, we could expect these results:

Time to build first car: 3 minutes
Time to build first batch of ten: 12 minutes

In the continuous flow example, the first car rolled off the assembly in in 3 minutes (18 minutes faster than the batch process).

All ten cars were built in 12 minutes (an improvement of 18 minutes over the batch process).

To really impress your guests, you could then demonstrate your worldly knowledge and command of automotive history by explaining that this is precisely what Toyota figured out when it began competing with Ford’s mass production system.

Toyota realized that a continuous one-piece flow process:
* increases productivity – more cars in less time
* builds in quality – easier to spot defective parts sooner
* is more efficient – less material lying around
* reduces costs – less inventory

By this time, other party goers will have no doubt picked up on your conversation, and you will most likely be the life of the party with everyone hinged on your every word.

You could excuse yourself.
You could change topics.

Or you could pickup a copy of Jeffrey Liker’s excellent book, The Toyota Way, and prepare yourself for the next party.

AddThis Social Bookmark Button


Tags: , , , , , ,

22 Responses to “Batch vs continuous flow processing”

  1. biff Says:

    Isn’t “continuous flow” another name for “assembly line”, which Ford invented?

  2. JR Says:

    Hi biff,

    In this context continuous flow refers to the continuous movement of cars from one stage to the next many cross functional teams working together on one car at a time.

    One of the many differences between batch processing and continuous flow is the amount of WIP (work in process). Batch process have a lot of WIP. Lean manufacturing tries to minimize WIP.


  3. JR Says:

    thanks to kingkongrevenge for highlighting another advantage of continuous flow:

    * is more robust – because continuous flow is so disruptive due to equipment failures, workers are forced to fix root cause problems fast

    This is another fascinating aspect of the Toyota Production System (TPS).
    I hope to write more about this later.

  4. crux_ from Says:

    This is wrong.

    In your batch example, the cars are finished with 30 person-minutes of labor.

    In your ‘continuous’ example… the cars are finished with 30 person-minutes of labor. The only reason that the time is shorter is because you have two more people working!!!

    Consider if you had three people working at each step of the batch process: It would only take 3 1/3 minutes to build 10 car bodies; 3 1/3 minutes to put on wheels, and 3 1/3 minutes to paint them. For a grand total of: 10 minutes! Batch processing for the win!

  5. JR Says:

    Hi Crux,

    That is an astute observation.

    In the batch processing example you have a lot of waiting (or waste).
    As machines are working, you have idle inventory sitting there, doing nothing.

    As you correctly pointed out, the continuous process doesn’t have that idle time. As soon as one product is finished, it is immediate sent to the next stage for processing.

    Regarding the other example you describe. It is not a similar comparison. In the example I use above, the assumption I use in both cases is that the machine can only work on one car at a time. Obviously if you bought x3 machine your would triple output.

  6. Chris Says:


    He is not wrong. In the example, the full teams to work paint and wheels sit idle until it is there turn to produce.

  7. crux_ from Says:

    I think there is a difference in assumptions here that needs to be made explicit:

    Chris mentioned that the “Full teams to work paint and wheels sit idle”. But I was assuming a resource (people) that could be moved between processes. Either imagine just one team who can build, put wheels on, and paint; or imagine that the teams “sitting idle” are not invited in that day and thus don’t draw down resources.

    In either case for a given amount of resources (people), and with the ability to allocate those resources between steps as you see fit, then the throughput will be the same.

    The example was designed to demonstrate that the “same” process is done more efficiently. But if you measure efficiency by person-hours, or dollars, in this type of thought experiment, you will find that they are identical. And if you have flexibility in where you put your resources, they are pretty much identical in terms of throughput, too. So touting the difference as one between processes is quite misleading, misleading enough that I would call it wrong.

    This is all academic, really. Nobody doing real manufacturing has “batch” and “continuous” bandwagons they cheer for: they use gigantic models, simulated annealing and genetic algorithms, actuarial accounting, economic modeling, and reams and reams of PhD-level research to carefully tweak everything to manage the complex interactions of throughput, risk, and demand.

  8. JR Says:

    Hi Crux,

    Yes you are correct. The throughput between the two system will be the same (all things being equal). The example was designed to compare and contrast the two systems as you said.

    We can agree to disagree on it being academic šŸ˜‰

  9. Greg Lins Says:

    Hey JR,

    I totally agree with you. Every time I discuss batch vs. continuous flow at a party (with or without cocktails) it engages the interest, fascination, and admiration of all attending. And it is much better if you can draw it on a napkin. Sometimes you can even tear up some napkins, label them, and move them around to show the differences. Awe-inspiring stuff….

    Anyway, I’ll toss in 2 more cents. Pretty much every implementation of continuous flow I’ve been engaged in results in inventory reductions of 50-90%, and floor space reductions of 30-70%. Defects drop off 90% or more, and labor efficiency increases 20-50%.

    What never ceases to amaze me is why so many businesses continue to manufacture the “old way.” My first work in this area was in the early 80’s, and you would think by now every production manager and engineer would understand the concepts. Go figure.

    Take care,


  10. JR Says:

    Thank you for the comment Greg and for sharing your experience in this area. Sounds like you have seen this type of thing before.

    I will look for you at the next party šŸ˜‰

    All the best – JR

  11. I Was Just so Wrong | The Hot Aisle Says:

    […] impress me? Both Toyota and the businesses I have been serving for the last decadeĀ specializeĀ in Continuous Batch Flow Processing. Toyota make cars in batch, Investment Banks like Lehman Brothers and Deutsche Bank process […]

  12. my blog Says:

    check this out…

    this is mine…

  13. negar Says:

    hi JR
    is ā€œcontinuous processesā€ another name for ā€œlarge scale processes”?
    and can i use this article for “batch and continuous processes” not flow?

  14. Bob Says:

    This doesn’t account for set-up time and other wastes.

    Imagine you have a ton of logs you need to move into your shed and stack them. You could pick-up a log, carry into the shed and stack – continuous flow. Or you could get your wheel barrow and shift in batch. Wheel barrow wins becuase there is less waste

  15. Deutsche Bank Says:

    Deutsche Bank…

    […]Batch vs continuous flow processing « Rasmusson Software Consulting[…]…

  16. Toyota Surabaya Says:

    Toyota Surabaya…

    […]Batch vs continuous flow processing « Rasmusson Software Consulting[…]…

  17. Sam Says:

    good comments abouth process flow,there`s cost reduction,space,inventory,transportation,scrap reduction etc.

  18. Work Flow Process Layout « Alicia Butler Pierre Says:

    […] was to reduce the time associated with application intake tasks (tasks 1, 2 and 4) by treating as a batch operation (specified times of day to handle incoming information) as opposed to continuous (handling as […]

  19. Marta Pastor Says:

    Claro y conciso. Muchas gracias. Besos!

  20. How To Reduce Waste, Increase Efficiency, And Boost Profits In Manufacture | Gearfuse Says:

    […] created a continuous flow process on his assembly line to speed things up. Now the cars came to the stationary workers. Moreover, […]

  21. Esta Figlar Says:

    Thanks man!

  22. Tony Li Says:

    Thanks for sharing the difference!
    When producing cars, batch mode wastes more time, but flow mode wastes more people, doesn’t it? For example, in the flow mode, it takes one minute to produce the body, and there must be nine other people to work at the same time to ensure to finish all ten car bodies in one minute; however, in the batch mode, only one person is required. Am I right?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: