This article is the third in our four-part series detailing how to deliver on-time, every time. Click here to read a holistic article which will walk you through the 4 steps you need to take – in the best sequence – to achieve high OTIF (on-time, in full). For the first step in more detail, have a read of ‘Release Dates are the Key to Reducing Lead-Times’.

Each article delivers value in its own right and can be read as a standalone piece.

In the previous articles, we discussed the importance of maintaining the right levels of Work-In-Process (WIP) through only releasing orders into the system at the correct time – obviously not releasing orders late, but also avoiding an early release. The aim should be to keep the WIP low and the flow fast through operations. We also highlighted the importance of aligning people on the priorities. Having a priority system which is clearly understood means that any orders that encounter a problem and become urgent, can progress quickly as a top priority. If your team is clear on the priorities, people will all be working on the right jobs in the right sequence; your capacity will not be diluted through multitasking. The WIP will be at the correct level, and you will ultimately achieve the right orders being completed on time – your delivery performance will shoot up to 80, 90 or even close to 100%. There is only one further challenge that has the potential to damage your high delivery performance and output… your bottleneck.

Your bottleneck determines your capacity and output. The number of parts that can be worked on in a day, a week, or a month is dependent on the performance of your bottleneck. By definition, a bottleneck is a work centre that systematically doesn’t have enough capacity to deal with the orders you want to put through it (we often must wait for the wine to pour through the neck of the bottle!). So, how do you identify your bottleneck? To find your bottleneck, look for the operation that has WIP piling up in front of it – queues can be a great indicator. Of course, if you haven’t taken steps one and two (controlling the WIP and managing the priorities) then most work centres will have a queue in front of them and your bottlenecks will be invisible amongst all the other queues. However, when the material release is restricted, the only work centres which should have WIP continuously piling up in front of them are resources that are capacity constrained. This is great! It shows you exactly where to focus. The place you should direct your improvement initiatives is the area that systematically cannot keep up with the demand you would like to put through it – it’s your bottleneck.

The first thing then, is that you must be able to see the queues. Visually on the shop floor, your operations and production resources and managers need to be able to spot the queues as they arise. Management attention should then be directed towards finding ways to increase capacity on the bottlenecks – these do not need to be complicated solutions. You now know the few key work centres you need to worry about, our bet is that if you go and look at them right now, they are not running at 100%. This tells you what you need to fix it. How can you increase the useful, productive uptime on those work centres by 5, 10, maybe 20%? If you can achieve that, the output of the whole facility would increase by the same amount…

There are some simple things you should try.

1. Make sure that the work centre doesn’t stop running while staff are changing shifts, having lunch breaks, attending meetings, etc. Organise cover so that the machine can keep running.
2. Offload jobs to other machines/stations where you can. It might be that another machine (a non-bottleneck) can do part of the process required, or it can complete the process but slower. It is worth setting up a parallel flow to ease the pressure on your bottleneck, even if the other work centres are deemed less effective, slower, and even if they are viewed as more ‘costly’ to run.
3. Use TPS, Lean and Six Sigma initiatives and techniques to reduce the setup time, downtime, waste and variation on these work centres (but only these key work centres). Immediately stop any TPS/Lean/Six Sigma work that is focused anywhere else and double the efforts on improving the work centres with queues.
4. If your overtime bill is high, this can certainly be reduced because you don’t want to sign off on any overtime for non-bottleneck areas. However, overtime on the bottlenecks should be the easiest financial decision you can make (assuming you have taken all the preceding steps, i.e., you have ensured uptime is high, work has been offloaded, setup times have been reduced and cycle times are optimised). Then, if you still have a legitimate queue, every extra hour of processing – according to these rules – will give you instant payback for any costs incurred manning those extra hours outside the normal working pattern.
5. Finally, create a ‘drum schedule’. The availability of a bottleneck work centre to produce parts determines the whole company’s success in making and delivering orders, we call this the drumbeat. The business operates in accordance with the beat of the drum. So, the bottleneck is the one place where finite, tightly controlled scheduling, and measuring the difference between what you should be achieving and what you are actually achieving, is massively useful. So, the last thing you need to be able to do is allow operators, shop floor managers, etc. to have tight control over these work centres’ performance. In other words, they should know exactly when each job is due to start, finish and how long the changeovers need to be. These should be measured carefully and controlled tightly. You don’t need this level of control anywhere else on the shop floor. Measuring it carefully will allow you to understand when you are hitting your target – which is great – and when you’re not hitting it, and why. When you don’t hit your target, you need to be able to capture data on the reasons why you couldn’t so that you can direct improvement initiatives towards making sure those causes are removed.

Taking these steps usually gives a huge jump in output (unless you are already maxed out). Your output shoots up; the jobs completed shoots up; invoicing shoots up; and your P&L will be looking really good. Once you have made these improvements, it might even be that your bottleneck will no longer be a bottleneck!

How do your information systems support this?

The first thing you must consider are the queues. Assuming you are controlling the release and clear on the priorities (see our other articles), you should be able to identify your bottlenecks and capacity constrained resources (CCR) easily. The way to do it is to have queues that are easily visible. Your ERP/MRP reporting software should be configured to visually show you at a glance where the queues are, and also the trends of the queues – are they increasing or decreasing? You can then always be aware of where your CCRs are and whether they are emerging or disappearing.

Next, you need to be able to produce finite schedules for those pieces of kit only. You should be able to choose the optimisation sequence, the timings, the setups, the run times, and so on to put together a highly challenging but achievable schedule that you can then measure variation against. So, the last thing your system should help you with is when you don’t hit the schedule (because things happen!) you need to be able to capture the reasons that prevented you from doing so. If you have a challenging but achievable schedule, and you fail to adhere to it, the reasons for why need to be properly understood. Furthermore, if those reasons repeat themselves, then they will be the next focus of your improvement initiatives – to remove the root causes. So, the three things that your supporting system should be able to do is firstly, to allow you to build a finite schedule for a key work centre. Secondly, recognise a delay in the finite schedule you have created, and thirdly, trigger a response to force staff to record the reason why.

So, if your systems can visually show you the queues emerging at work centres, help you to produce finite schedules for key pieces of kit, measure variation away from that schedule and force you to capture good root cause reasons when the schedule isn’t hit – you’re in good shape!

If your system lacks any of these functions, you may wish to contact us for more information on Ropeweaver. Ropeweaver is a tool designed to support companies’ native ERP/MRP systems. Click here for an article detailing all the areas it can help you with, or get in touch for a demonstration/more information.

To be the first to know when the final instalment of our ‘High OTIF’ series is published, you can follow our LinkedIn.

_{By Phil Snelgrove, Lauren Wiles}

© 2022, Goldratt UK