Powerful demonstration. Thank you. Would love to see more of these examples and philosophies demonstrated beyond the production realm and 'in the office'. Great job Ryan.
For years, I have worked to shift our production towards a one-piece flow approach. However, certain processes with setup times exceeding five minutes-such as configuring a timber moulder for a specific profile-diminish the benefits of one-piece flow. Products with only a few, same-sized components are also better suited to batching. Despite these limitations, we can still incorporate elements of one-piece flow into our assembly process. For example, packing corrugated boxes one at a time as products are completed reduces the need for stacking and improves efficiency. Even so, I have only managed to achieve a one-piece flow for about 30% of our processes.
This is where SMED (Single Minute Exchange of Die) comes in: continuously improve your setup process so you can change setups as quickly as possible. If you were able to configure the timer moulder in 1 minute instead of 5, would that allow you to reduce the batch size further? Another opportunity is having multiple dedicated machines for a specific setup. If you had an additional timber moulder with a different profile ready, then changing setups would be as simple as choosing which moulder to use. A much simpler and more attainable example is having several hand routers with a dedicated bit for each operation, so the tool setup never has to be changed. You're correct that there are eventually limits, but each might require a new type of improvement to overcome.
This demonstration worked because each step required zero setup. For many manufacturing steps, this would fail horribly. If my employees took one item and set up the paint sprayer and loaded it with paint and then painted ONE thing and moved to the next step, we’d fail immediately. The paint sprayer would have to be cleaned out and we’d lose tons of paint. I like the idea here, but sometimes these principles seem presented as a one size fits all, and there is a lot of nuance.
single-piece flow is an ideal to strive towards. Just like "no defects" and "no waste" However, the ideal of "no defects" and "no waste" only gets possible when you've implemented the smallest batch possible as this drives the least amount of variables impacting the creation of the product. It also drives the fastest feedback look possible as a small batch can be shipped to Customer as quick as possible, and wtih that feedback we avoid reproducing issues that existed with that initial batch or even add improvements. variability hides and produces defects and waste. moving towards single-piece flow smokes up issues before they become fires i think another way to approach the principle of "single-piece-flow" is not see it as fuzzy or nuanced but as math highlighting variation. The more variation we have, the more issues we have. Somethings seem impossible, at times nonsensical, but with the approach of math ( count of variables ) determing quality and speed to the Customer we can tap into our imaginations and Front-Lines in getting creative and as we strive to see how close we can get to ideal.
You are right. In cases the workloads between different flow steps cannot be levelled, the difference has to be compensate by a buffer or by a batch production and combined or not with different work times. It is a required waste to secure flow while you don't have better conditions to do the best. As mentioned by @JonSchleicher, one-piece flow is the lighthouse to move towards to, not necessarily the possible current state. Eventually, setup times have to be reduced, step-by-step, to allow better synchronization, one-piece flow and flexibility.
This presentation was flawed. In the batch version 2 people were not employed during the whole process, of course if you have them later on all employed that the customer will receive items faster when 3 people were working for him at any given moment and not just one working while others are waiting. In real life when you are doing batch work each station always has a job to do, the one who starts the order first is the one who begins the next order first. One piece flow can have better efficiency but this was flawed example of showing it.
It is not a completely fair comparison indeed, but it is still a valid demonstration of the benefits of one piece flow. To make it even better though, I would suggest to load all processes steps with a full batch and then measure the time to deliver a NEW order coming from the customer. The throughput time for the batch production would be the same or even worse than the one from the video, if the downstream process person is slower than the upstream one (in other words, if the next process is not finished processing the previous batch when the new batch of 10 dices arrives). This could also be used to demonstrate the substantial difference in production flexibility when the customer demand changes last-minute (in batch production the batch is in the middle of the flow and has to be fully reworked before moving on, whereas in one-piece-flow it gets way less troublesome to adjust).
Love the dice example. That's a keeper.
Powerful demonstration. Thank you. Would love to see more of these examples and philosophies demonstrated beyond the production realm and 'in the office'.
Great job Ryan.
Great illustration. Thanks Ryan
Great illustration! Yet so simple! Thanks, Ryan
Glad it was helpful!
Ryan, this video is the best example of one piece flow. I saw and shared many videos related, but this one is super clear, simple and on pont!!!!
Thanks
For years, I have worked to shift our production towards a one-piece flow approach. However, certain processes with setup times exceeding five minutes-such as configuring a timber moulder for a specific profile-diminish the benefits of one-piece flow. Products with only a few, same-sized components are also better suited to batching.
Despite these limitations, we can still incorporate elements of one-piece flow into our assembly process. For example, packing corrugated boxes one at a time as products are completed reduces the need for stacking and improves efficiency. Even so, I have only managed to achieve a one-piece flow for about 30% of our processes.
This is where SMED (Single Minute Exchange of Die) comes in: continuously improve your setup process so you can change setups as quickly as possible. If you were able to configure the timer moulder in 1 minute instead of 5, would that allow you to reduce the batch size further?
Another opportunity is having multiple dedicated machines for a specific setup. If you had an additional timber moulder with a different profile ready, then changing setups would be as simple as choosing which moulder to use.
A much simpler and more attainable example is having several hand routers with a dedicated bit for each operation, so the tool setup never has to be changed.
You're correct that there are eventually limits, but each might require a new type of improvement to overcome.
HUGE !!
Tremendous video!
This demonstration worked because each step required zero setup. For many manufacturing steps, this would fail horribly.
If my employees took one item and set up the paint sprayer and loaded it with paint and then painted ONE thing and moved to the next step, we’d fail immediately.
The paint sprayer would have to be cleaned out and we’d lose tons of paint.
I like the idea here, but sometimes these principles seem presented as a one size fits all, and there is a lot of nuance.
single-piece flow is an ideal to strive towards. Just like "no defects" and "no waste"
However, the ideal of "no defects" and "no waste" only gets possible when you've implemented the smallest batch possible as this drives the least amount of variables impacting the creation of the product. It also drives the fastest feedback look possible as a small batch can be shipped to Customer as quick as possible, and wtih that feedback we avoid reproducing issues that existed with that initial batch or even add improvements.
variability hides and produces defects and waste. moving towards single-piece flow smokes up issues before they become fires
i think another way to approach the principle of "single-piece-flow" is not see it as fuzzy or nuanced but as math highlighting variation. The more variation we have, the more issues we have. Somethings seem impossible, at times nonsensical, but with the approach of math ( count of variables ) determing quality and speed to the Customer we can tap into our imaginations and Front-Lines in getting creative and as we strive to see how close we can get to ideal.
You are right. In cases the workloads between different flow steps cannot be levelled, the difference has to be compensate by a buffer or by a batch production and combined or not with different work times. It is a required waste to secure flow while you don't have better conditions to do the best. As mentioned by @JonSchleicher, one-piece flow is the lighthouse to move towards to, not necessarily the possible current state. Eventually, setup times have to be reduced, step-by-step, to allow better synchronization, one-piece flow and flexibility.
This presentation was flawed. In the batch version 2 people were not employed during the whole process, of course if you have them later on all employed that the customer will receive items faster when 3 people were working for him at any given moment and not just one working while others are waiting.
In real life when you are doing batch work each station always has a job to do, the one who starts the order first is the one who begins the next order first.
One piece flow can have better efficiency but this was flawed example of showing it.
It is not a completely fair comparison indeed, but it is still a valid demonstration of the benefits of one piece flow.
To make it even better though, I would suggest to load all processes steps with a full batch and then measure the time to deliver a NEW order coming from the customer. The throughput time for the batch production would be the same or even worse than the one from the video, if the downstream process person is slower than the upstream one (in other words, if the next process is not finished processing the previous batch when the new batch of 10 dices arrives).
This could also be used to demonstrate the substantial difference in production flexibility when the customer demand changes last-minute (in batch production the batch is in the middle of the flow and has to be fully reworked before moving on, whereas in one-piece-flow it gets way less troublesome to adjust).