— January 10, 2018
We’ve all heard about the importance of focusing on one task at a time.
Take that idea, apply it to manufacturing, and you have the concept of one piece flow.
The concept is simple in that you have to limit yourself to working on one item at a time (you only ever have one WIP item). By doing this you can save time and money, as work is performed faster and any space you use is highly optimized.
In fact, it’s the very same concept that Toyota used to reduce the time taken to pack boxes in to help families hit by Hurricane Sandy by 94%. In total, the aid workers were able to feed 400 more families in less than half the time.
However, there are a few vital things that you need to be careful of when applying one piece (or “continuous”) flow. It’s not a miracle cure, and while it can be used outside of an assembly line, there are a few practical elements which can’t be ignored.
So, today I’ll be diving into:
- What one piece flow is
- The advantages and disadvantages
- How to implement it
- Using one piece flow in any area of your business
- The problems of theory vs practical effects
Let’s get started.
What is one piece flow?
“One piece flow” is all about reducing WIP to the point where everything is either waiting to be started, in progress, or complete. That’s why it’s also called “single piece flow” and “continuous flow”, as everything is constantly progressing and only one item is in any given queue at a time.
That means that there is a maximum of one item at any given stage of the process. Multiple items can be in the process as a whole, but nothing has to queue up and take up space from the moment it starts until the moment it’s complete.
This reduces waste (you’re not spending money holding items for future delivery) but requires a huge amount of coordination and consistency to pull off.
There’s also a big difference between the theory and practical application of this model which makes the whole concept a little muddy, but we’ll get to that later.
By limiting the number of items in your queue to one you’re naturally reducing the time any given item spends in the system. This can be shown through using Little’s Law, since your lead time should decrease proportionally to your WIP.
For example, let’s say that you need to pack boxes for a supermarket delivery. Every box contains the same 10 tins and 10 bottles, and you have two separate workstations, one for packing the tins, and a second for packing the bottles.
The slower method would be to batch the packing together so that all tins for all boxes are packed before moving the batch on to the bottle station. This means that only one box is being filled at any given time while the other nine are left in a queue.
However, one piece flow would be different. Rather than waiting for all tins to be packed, boxes would be handed to the next workstation as soon as they were completed. This means that (assuming the packing time for bottles and tins is equal) two boxes are being worked on at any given time and none are left in a queue.
You’re almost doubling the work being done at any given moment, thus greatly reducing the time it takes to complete a whole batch.
While you need a solid business process management strategy in order to make sure you’re taking a consistent approach, the benefits of using this model in the appropriate circumstances are massive. You’ll be saving time and money (in manpower, real estate, and little to no unused inventory), boosting morale, increasing the quality of your results, and much, much more.
In fact, let’s run over the pros and cons right now.
There are two main advantages of one piece flow; that your work is completed faster and that you save money both due to the time and space you free up. That said, these simple benefits have huge knock-on effects.
First up, using one piece flow can make your final results higher quality than if you tackled them in bulk. This is because errors are more likely to be spotted partway through the process than at the very end when it’s too late to correct them.
Mistakes can also be spotted mid-batch before they affect the entire order.
For example, if you assemble 15 computers at once and use the wrong type of transistor, that error will be present in all 15. Making them one by one can save you that pain by having the first faulty computer detected and the mistake corrected before any others are assembled.
Next, your processes will be more flexible to the customer. Since you’re able to fulfill orders and tasks much faster, you can leave them for longer before starting work on them. This then leaves more time for the customer to make any alterations before you’ve already started work.
Not to mention that the extra time and money you save make it easier to scale up your operations and do so reliably, since you can monitor the production of each product across every step of the process.
Depending on how your process is set up, continuous flow can also help to increase the morale of your workforce. As long as there is still some variation in duties (eg, switching employee tasks every so often) the higher rate of progress can help to show your employees that the work they’re doing is making something worthwhile.
Finally, by breaking up your process into set workstations and blocks of tasks you can more easily see where you can benefit from using business process automation. Any initial cost of setting up the automation would quickly be offset by both the money you save with one piece flow and with the savings your new automation will make going forward.
The main problem with one piece flow is more down to how it’s used, rather than the method itself. This model isn’t suitable for all operations, and attempting to apply it to everything will do far more harm than good (again, I’ll cover this in more depth during the “theory vs practical” section further down).
For example, continuous flow thrives where you’re producing a very small variety of products. Batching together several different items from an order that require different approaches will result in confusion and inefficiency.
Not only that, but if your products or batches are radically different then it can be expensive to change your production lines to suit each order. While this isn’t a problem if you’re only making one product or performing one process, the downtime to set up your equipment means that the unit cost can end up being greater than if you had batched everything together.
Thus we come to the next big disadvantage; that it’s hard to implement and very easy to get wrong. This is mostly down to how you divide up the work between stations and the number of employees required to make each as efficient as possible.
For example, in the video below we can see some proof of how much faster one piece flow can be than batch processing. However, the argument can be made that the process could be sped up even further by separating the work into different stations, having one person fold the boxes, one put the stickers on the boards, and so on.
Finally, while continuous flow makes your processes faster and (sometimes) cheaper to run, there is a danger that employees will lose motivation if they’re doing the same monotonous task for too long. In turn, this can impact the quality of the final product – even if mistakes are caught soon after they’re made, the whole process needs to be stopped and the source of the error found in order to make sure it doesn’t keep happening.
Imagine working on a production line, screwing the tops onto toothpaste tubes all day every day. Technically the process is efficient in terms of time and money, but if there’s no variety in your then you’ll soon go stir crazy and zone out, leading to mistakes.
The solution to this is admittedly to automate whatever simple, monotonous tasks you can, but this can be difficult. You may not have the resources required to set up the automation, or the space to house the required machinery, and so on, so sometimes the only solution is to rotate employee duties (or let motivation drop).
How to implement continuous flow
Now that we’ve got the basics out of the way, it’s time to dive into how you can implement continuous flow in your own operations. This mainly applies to manufacturing, but (as I’ll cover below) these same principles can be applied further with a little tweaking.
Meet the requirements
Since one piece flow is such a wide-ranging concept, there is less of a go-to guide for how to apply it. Instead, by meeting the following requirements you can change your operations to follow a continuous flow principle.
- Limit your WIP to one item in any given queue
- Keep your uptime as close to 100% as possible
- Divide work and resources evenly between workstations to prevent bottlenecks
- Measure and be able to consistently repeat the time taken to complete any given step
- Scale your process and operations to demand
Many of these factors are related to each other (eg, keeping high uptime prevents a bottleneck or queue forming) and so will fall in line quickly if addressed as a whole. Sure, it won’t be easy, but knowing what you need to do and assessing what’s the easiest problem to solve first is half the battle.
It’s easy to become lost in the specifics, but as long as you’re measuring aspects such as the time taken to perform a given task, the performance of a particular machine, person, or process, and are able to scale your operation to meet demand as it rises and falls, you’ll be able to build an efficient, cheap, and predictable business.
Take a systematic approach to your processes
If you’re a long-time reader (or you’ve noticed our company’s name – Process Street) you’ll know that we’re all about how important and powerful it can be when you document workflows and improve your processes.
Usually, it comes as an addition to the topic we’re talking about, such as saying how your operations can be better if you document them properly, and so on. This time it’s core to the entire theory.
Continuous flow is impossible to achieve without documented processes.
In order to break up the required work you need to be able to see how everything is achieved, how long it takes, and how skilled the work is. Only by knowing all of this can you effectively break up your process into stages to successfully limit your WIP.
Plus, this comes with all of the usual process documentation benefits, making it much easier to spot gaps and mistakes, optimize outdated or inefficient sections, automate what you can to save time and money, make your employees more accountable and motivated, and much, much more.
Analyze and optimize your operations
Finally, we come to actually analyzing and optimizing your operations. This is done by enacting the following steps:
- Choose a process and make sure it’s suitable for continuous flow
- Divide the work involved into separate blocks (“workstations” or “work cells”)
- Make sure you can meet the required production time
- Create a floor layout
- Balance your cells
- Audit and improve your flow and processes regularly
Once again, while this is simple on paper you’ll need to be careful of the practical elements involved in its application.
When separating your work into blocks, you’ll generally want to group tasks by their type or the equipment involved. However, this can easily result in some blocks containing more work than others, thus forming a bottleneck.
Your work doesn’t have to be completed in the same place, and there lies the solution to your bottlenecks. Move tasks that don’t need to be completed in a set place or with specific expertise (eg, simple assembly) to a different, smaller block in order to maintain a steady product flow and an even workload across your cells.
When planning the floor layout for your employees and machinery it’s also important to consider the travel time between workstations. If possible, the most efficient way to set up your process your be to have a “U” shaped setup, as this minimizes both the travel time between stations and the space that the process takes up.
Using one piece flow in any area of your business
While there are many issues with applying one piece flow to a practical setting the idea is undoubtedly solid – in this way it’s easy to see why one piece flow is so closely associated with lean. If other factors can be regulated and maintained then reducing your WIP items (waste) will make your operation run quicker by default.
The benefits don’t have to stop with just manufacturing though.
Continuous flow can also be applied to settings outside of manufacturing, such as content creation, CRM, and human resources. As long as all factors are considered and a balance is struck, there are a huge number of benefits that can be had by applying this idea to even small sections of your regular documented processes.
The key is not following the theory so doggedly that you lose sight of what small tweaks could do to improve the practice given the context you’re bringing it into.
Let’s say that you want to bring continuous flow to your content creation team.
You could apply the theory to the editing process for your blog posts quite easily. Rather than reading through three posts to check for spelling, then reading through them for grammar, then again for structure, and so on, you’d complete the task much faster (and avoid going insane) by completely reviewing a single post before moving on.
To take a real-world example, over at Buffer, Kevan Lee wrote a fantastic post about the process he used to write posts within three hours on average. Aside from being a great read, his method (seen in the image above) is a kind of compromise on continuous flow which factors in the mental exhaustion of any one of the three blocks to writing a post.
Theory vs practical
There is a very real danger with one piece flow that I’ve mentioned already – that the practical application of it will completely fail if not considered and enacted carefully. This is because the theory is so simple that it’s easy to forget the hundreds of other elements that can influence it.
Seriously, I’ve already scrapped this post once and started from the ground-up. It’s all too easy to get lost in the complications.
One problem is scope, as if you get it wrong then continuous flow will not be suitable to use.
This is because almost every process is both continuous and batched depending on the level you’re assessing. Manufacturing, for example, is often done with continuous flow, but if you were to include delivery then suddenly you’re also working with batches.
Speaking of which, you can’t take for granted that one piece flow is the best thing to do. Doing so in practical terms will lead you to waste huge amounts of time and money in the pursuit of a theory, rather than the benefit it’s intended to bring.
Sticking with our delivery example, it makes absolutely no sense to deliver one item at a time (unless the item is huge). While this would be the ultimate in continuous flow, practical elements such as the travel time and cost, along with the delivery having a set start time mean that it is far more efficient to increase your batch size.
Motivation can also be a massive problem, as the variety that’s required in order to keep employees happy doesn’t serve continuous flow on paper. Swapping employees between tasks makes them less experienced and effective at any one item, and extra training might be required as opposed to just keeping everyone on one duty.
Having said that, if your orders allow it you could make up for these issues by having continuous flow until the orders are complete, then switching your workforce to another task. While this is more applicable to small teams than large manufacturing plants, you get the idea – like all things, one piece flow is best in moderation, and with careful consideration.
What do you think of one piece flow or batch production? Have any horror stories of lean or one piece flow gone wrong? I’d love to hear from you in the comments!