The only difference I see between one piece flow and batching is that in one piece flow, you avoid have to pick up the same object multiple times – e.g. if you washed all dishes first then dried them, you would need to put the washed dishes down before washing the next one, and then picking the same dishes to be dried. While in one piece flow, the dish could just be dried before you put down.

I am still struggling to understand how else is one piece flow better than batching. Note that in the subway example, all the necessary ingredients of the sandwich had some form of preparation beforehand – lettuce and tomatoes sliced, slices of ham portions

There are instances where one piece flow doesn’t make sense – for example in tools that are designed to “work on” several pieces at a time or job typing process – e.g. you need to set up your machine for each different product you are producing and the set up time is very long. Going with the lean and one piece flow theory, you only manufacture when you have a demand for your product, but this in reality might cause you to waste time to constantly set up your machine while your operators sit idle waiting for the set up to be done. For example going with the subway example, you will be handling the tomato and lettuce many times if you were to cut them when an order comes through, instead of slicing them all at once.

Great comments, Joel. I agree about the size of the batch. I often tell people to focus on reducing batch sizes as much as possible, even if you can’t get down to one piece.

doing one piece vs batch queue..who knows batch and queue may maybe faster.
it’s really case to case basis and the time study needed to be done unbiasly
repeatedly to verify actual result. And it really depends on the batch size:

– Like getting 6 cheese one time and distribute may save time in reaching for cheese 5x less. (bread, etc.).
– Batch and queue may not be effective if there’s extra motion to move (reach or back and forth movement of your WIP)

So if you primary aid is speed, then go ahead and do time study which is better.
other advantages of one piece flow is more important: then do one piece flow

quality: (especially if two or more people)
– less risk in skipping a step (like you can easily get interrupted, for one piece flow you can finish the process before stopping)
– identify problem faster
– consistency
– if this is sensitive part that can be damaged you do not want many parts on one table, as bump each other

Others:
– save space
– less risk in rework/scrap (as less inventory) after detecting an issue
– if a safety issue
– if next customer is waiting, you do not want a big lot size.
– more flexible to change order or setup

If you want benefit of both maybe do hybrid or smaller lot size.

If two persons or more then this debate is less and one piece flow will be more effective in speed and quality.

Thank you.

Imran,

Excellent point.

Thanks,
-Imran.

Also to add to my comments if you are in a food service and you get bulk orders, you will probably increase your work stations or will have a different production line for bulk orders where in you will have one person preparing the breads, one spreading the mayo and butter, then another one placing your ham and another one placing the lettuce and pickles, then another person wrapping the sandwiches. Tis isnt batch and queue, its an assembly line

I agree with steve, Part of the process is set up time , in case of food preparation, you will find that all vegies have been sliced and laid down on a cooler ( set up), take the caseof subway sandwich , when you order, they prepare sandwiches FIFO, they finish one sandwich at a time. If you want the line to move faster , you can add stations. If you know that you,will be preparing many sandwicheson demand nit on schedule, you will be cutting your vegies ahead of time and have an assembly of these vegies ready to be placed in every sandwich ordered.

Everyone thinks batch flow is faster than one piece flow; UNTIL, you actually put them to the test. My wife washed dishes with a half one piece flow and half batch method. She would wash and rinse a dish then set it aside, wash & rinse the next dish and set it aside. She would continue this until all dishes were clean. Then, she would dry and put away a dish one at a time.
I told her she should try washing, rinsing, drying and putting away a single dish and then move to the next one. We timed both methods with one piece flow being 30 seconds faster – given we did 10 dishes as a sample.

Making sandwiches, you have to set all the supplies out in an assembly line setup to get an accurate timing for one piece flow. No one in there right mind is going to go from the counter to the fridge, get one slice out, walk back to the counter, put it on the sandwich. No. You will have all the meat set out, all the cheese, toppings whatever in an assembly line. I haven’t tested it yet; but, I’ll just bet it will be batch flow hands down.

In the PB&J example – instead of 10 sandwiches now there is 1000. Assume you can get paid as soon as a sandwich is completed. Also assume that the customer may want grape, strawberry, or rasberry jam. A one piece flow scenario easily adapts to different market requirements. A batch method is not flexible in fact if the customer or the company changed their minds halfway through the process (Engineering Change Notice) then a great deal of pieces of bread with the wrong jam spread on them will be tossed, along with the labor it took to apply the jam.

The problem with simplistic examples is that they seldom capture the true nature of what is really valuable in terms of processing methodology. Flexibility should not be overlooked, neither should the amount of inventory on hand.

Back to one-piece-flow. I would imagine that if you have more than one operator it would be more efficient to do one-piece flow. You may have to do the calculation prior to make it cost-effective. One-piece flow works.

Seriously though, the PB&J Batch vs Flow manufacturing argument needs to take account of (i) Setup times and (ii) Process Times.

To make Flow happen you have to minimise setup time, typically via dedicated lines or modular jigs etc. In this case setup time is from knife changeover (assuming you wash the knife between PB and J operations). Using 2 dedicated knives clearly moves the cost per sandwich up to uneconomic levels.

Process time is a tradeoff between minimum time to complete (typically better in batches) vs other costs (inventory & WiP storage, defect control etc). In this case it is my observation that the on-cost of extra BiP (Bread in Process) in front of the workcentre was negligible and SiT (Sandwich in Time) delivery was still satisfied, so the Batch Process is probably optimal here.

However, if you have to manufacture a variety of products on the same chassis the issue become one of relative homogeneity. Clearly you have some easy variety options – you do not need to be Fordist and offer any jam so long as it is strawberry – the Toyota approach would recommend you using multiple Jams, scheduled to user demand forecasts, on the same braed and PB subassembly.

Stick control here could be via 2 bin rather than JiT delivery, saving costs I think.

However to offer the Alfred Sloan approach and offer 5 variants – caviar, pate, chocolate spread, PB&J, and butter – will require a more complex setup and negotiation of raw material delivery, thus necessitating ensuring higher values at point of sale.

Bad Batch-and-Queue:

To build sandwiches the Batch-and-Queue style I would take 3 pieces of bread (3 lunches to make) and place peanut butter on it. This work actually has to be done by your daughter, because she is part of the Peanut Butter union. This This would be done down in the basement. I would have to call my wife to come get the bread at some point (whenever she had time).

The bread/peanut butter pieces would then be carried to the garage to sit and wait.

Then, let the 3 pieces of bread with peanut butter sit, while I take 3 pieces of bread and place jelly on it. This work would be done by your son, who is in the Jelly Union. This would be done upstairs in the baby’s room. I would call my wife to come get the bread/jelly when it was convenient. We’d make 4 pieces actually, to account for damage that might occur when moving the bread/jelly. The bread/jelly would be moved to the garage some hours down the line.

We’d gather in the kitchen. People want 2 sandwiches, but it’s more efficient to 3 sandwiches at time (to keep our cost down). Upon deciding to build 3 sandwiches, we would have my wife go to the garage to get 3 bread/peanut butters and 3 bread/jellies.

Then, I’d marry the two sides of the bread and be done with 3 sandwiches. I have to do this work myself, being a certified “Assembler” job class.

One sandwich would get thrown away after growing moldy.

Now THAT is a better (and not unrealistic) batch and queue example. And, no, I’m not blaming unions for all that is inefficient in the world.