Dealing With Bottlenecks
Before you buy a new piece of equipment because you think you have a bottleneck, use a machine balance chart so that you know the real score.
Balance Charts
During the course of value stream mapping, you will probably use balance charts to understand the nature of your processes, and to understand where potential bottlenecks will occur. The balance chart is a powerful tool that gives you enormous insight into your value streams. Balance charts can be drawn for the work the operators do, or for the processes in a value stream, and even for effective machine cycle times.
It is rare that you will find the operators to be a bottleneck in a process. After all, it is a relatively easy thing to rebalance the work elements, or to add or move operators if this were the case. But it might be quite common to find a machine that is a bottleneck in the value stream. Adding and moving machines is not always an easy thing to do - often requiring capital investment or operating investment to do so.
A bottleneck is any machine that cannot produce fast enough to meet the required takt time. ('Takt' is the rate of customer demand). To find out whether a machine is in fact a bottleneck, we need to determine the 'effective machine cycle'. The effective machine cycle time is the machine cycle time per piece, plus the load and unload time (during which the machine cannot cycle), plus the changeover time divided by the number of pieces between changeovers. Once we have the effective machine cycle calculated for each of the machines in the process or value stream, we can draw a machine balance chart.
Examining the balance chart above gives us several insights as to what is happening. First, we can see that machine number two 'M2' is a bottleneck because its effective cycle time is greater than the takt time. It cannot meet the rate of customer demand, and most companies will be running this machine on additional shifts, or operating it on overtime to keep up. In our travels, we quite often see bake ovens and paint booths working additional hours to keep up with production.
We also notice that machine number 4 cycles close to the takt time, which may present problems in the future if our product grows. (And our takt time line drops)
Machines one and three seem to have enough capacity for now.
One of the key strategies of a lean approach is to mix production at the lowest level possible to take inventory off the floor, reduce lead times, increase flexibility, and to become more responsive to defects. (Lowest level means making all of the possible product configurations in the product family within a very small timeframe). If we were to mix at a lower level in this case, the bottleneck problem may become much worse unless changeover times are reduced dramatically. (There would be more changeovers in our calculation for the effective machine cycle)
What do we do?
First of all, make sure that no defects go through the bottleneck. Every time we process a defect at a bottleneck we wasted valuable time that could have been spent processing a good piece.
Reducing changeover times will bring down our effective machine cycle time - and possibly help us to achieve our takt time.
Continuously improve the load, start, and unload times of the machine. Doing this will also reduce our effective machine cycle time.
Eliminate waste in the machine itself. For example, if we were drilling items of a one-inch thickness, then why would we let the drill bit move ten inches on each stroke? Reducing the distance traveled will reduce the cycle time.
We could split apart some of the things that the machine does and have them performed on another machine.
We could use two machines and alternate between them.
We could use several footprints of the same work center - thereby increasing the takt time line.
We could run our cell on an additional shift to increase takt time - but this comes with additional overhead expenses that would be associated with running another shift.
Finally, we could decouple the bottleneck from our workflow, and operate it using a supermarket pull system.
Conclusion
It's actually amusing how many times we've been on the plant floor of companies that want to show us their bottleneck operation. When we actually stand and watch the operation we are surprised (not really) to see how often the machine is not actually producing parts. The operators have gone for lunch or break, and the machine is stopped! They are away searching for equipment or supplies. The machine is broken and the operators can do nothing while a technician makes the repairs.
Bottlenecks should be managed as valuable resources. Every minute counts. So good planning, and a preventative maintenance program are essential ingredients to achieving takt time.
The hardest part to manage is the last resort - decoupling the bottleneck operation from the rest of the process flow using a supermarket pull system. We are currently writing a workbook on advanced flow that will deal with this topic. It will be available in the near future.
So our advice is that before you consider buying a new expensive piece of equipment because you think that you have a bottleneck, please first use a machine balance chart so that you know the real score.