Improving Flow with Changeovers

One of the 5 fundamental principles of lean is to obtain flow, find a way that your product or service flows seamlessly from the start of your process to your client and do this as fast as is practicable to maintain a high-quality product or service. There are always many ways to do this and it will vary from organisation to organisation since every organisation, even ones in the same market are different, they have different people, equipment, raw materials, suppliers and of course customers. So, the question is where do you start if you want to increase flow? Unfortunately, there is no one right answer, it depends on you!

Batch Size Reduction 

Many people start by jumping in and reducing batch sizes, after all both lean and Theory Of Constraints (TOC) advocate for smaller batch sizes. The typical rule of thumb is that is you half your batch size, ultimately you will increase your throughput to around double what it was. You will also cycle through the process of making all your parts more frequently increasing your EPEI (Every Part Every Interval) measure. That means your stores' inventory will reduce, free up cash and reduce the liability of the stock sitting on your shelf and yes, I said that stock is a liability, irrespective of what your accountant may be telling you. Stock sitting in your warehouse on the shelf is a liability and costing you money, it's money tied up, you may even have had to finance it in some way so you will be paying interest on that money and what if the market disappears or your competitor launches a product twice as good at half the price, is that stock on the shelf still worth the value it has? So, reducing your batch size makes sense, right? Well yes, but it should not be your 1^st point of attack, if you just half your batch size but your change over times remain the same then you will kill your capacity, quickly.

Change Overs 

When we start thinking about the wastes in our process and really looking at it and really seeing things for what they are, lots of non-value adding work getting in the way of flowing your product, it's not unusual to decide to attack everything to eliminate any waste in them very quickly, but this scattergun approach very rarely works. So where do you start?

Changeovers are typically one of the largest areas of obvious waste and tend to be the thing that will inhibit your batch size reductions as well. If you have a 5 or 10 hour change over you are unlikely to want to do that too often so you build as much as 'economically possible' (which is an oxymoron!) which drives up the batch size and the stores' inventory. So by starting with your changeovers you can earn your way to smaller batch sizes, for example if you can half the time for the changeover then you can realistically half the batch size with no real impact on your capacity (even though you are building more than you really need which is what is stealing your capacity).

SMED – Single Minute Exchange of Die

People sometimes get confused when we start talking about single minute exchange of die (SMED) changeovers, they say there is no way they can change a machine or a process in a minute and that's fair, and really it should be single digit rather than minute, the aim is to be able to change over any product or process in less than 10 minutes. Just think what that would be like, any process in under 10 minutes, you would have no reason not to change on a whim, to run as small a batch size as you could right? The push back you normally get is that the machine is a pain to change, people aren't good at it, the answer is always OK, how do we simplify it and people will never be good at it unless they practice and do more.

The process of a SMED improvement isn't all that complicated, well let me rephrase that, it can be as complicated as you want it to be. You could get your stopwatches out and monitor a whole sequence step by step then chart it, analyse it and hold meetings about it or you could talk with the people working on it and start there. My favourite question is always, tell me the thing that annoys you most about changing over this machine. Quickly we get big-ticket items of where things are, tool positions, having to manhandle something instead of having a lifting or sliding device, maybe you actually need 2 people to change the machine over better. Perhaps it's the fact that the tooling or set up for the next job isn't ready and they have to go look for that. Starting the SMED process with a simple 5s workshop is always a great step, you instantly clear the clutter and get the area laid out well, ensuring you have the tools that you need immediately to hand, the instructions available to hand, limit the bending and twisting the people have to do. Having an area set for the next job to be waiting in makes it easy to see that it's not there and some training around not starting a change over unless the next job is there, and it's been fully checked is always good.

You can make dramatic impacts on your changeovers just by taking basic steps like these and the figuring out the best sequence of operations even before whipping out the stopwatch and the chequebook, in fact, the chequebook (do they still exist?) should be the very last option, is there something you can do in house to simplify it? Following the lean mantra of easier, better, faster, cheaper (in that order) take you a long way, of course overriding everything there is it must be 100% safe at all times.

The more you study your set up the more you will identify the wastes that inherent there and it's not unusual to see all 8 wastes in the set-up process. This is where to focus your initial efforts when it comes to shrinking batch sizes for better flow.

Summary 

If you want to improve your flow then there are lots of things you can do, review your layout for a start, review the wastes but ultimately you are going to come back to batch size which means you need to fix your set ups and reduce them as low as possible. Reducing batch sizes without focusing first on set ups is frankly foolish and will steal capacity rather than improving flow. When you start looking at the changes keep it simple, get the low hanging fruit and get the team involved from the start. Only once you have gotten the set up time down by at least 25 – 50% think about your batch sizes.