In my previous article I presented what are the bottlenecks and the Theory of Constraints proposed by Elyah Goldratt in the 20th century. As a reminder, the bottleneck is a factor that limits the system's throughput. As much as the bottleneck is possible to do - that's how much the whole plant can do.
The plant in which everyone works is highly inefficient.
Imagine that you are going to your partner's production plant. You see that two of his subordinates have a great time watching youtube. They eat crisps and have a lively conversation about the last match of the Premier League. On top of everything, these employees seeing your partner do not hide behind the machine - they shout "good morning" with a smile and nod. What do you think in this situation? "Maybe they have some tapes on him?" However, I hope that thanks to careful reading of the first part of the article, you perfectly understand the situation here, and your association falls on "he knows what he is doing". As I mentioned earlier, the factor that is not a bottleneck is by definition a bit of a "slack". Exploiting this play has a bad effect on the company's financial situation, because it freezes its financial resources in inventories.
"So, should I let my people not work?" "I'll raise costs by that!"
If these are employees who do not work directly on the bottleneck, then the answer is YES. Let them sometimes do nothing. Well, unless you decide to give their wonderful job cleaning the plant - it is a relatively frequent treatment found on "warehouses". The useful use of such a subordinate may be a longer time for him to check the quality of the semi-finished products he made. You can also make a small buffer (several days), which in the event of a machine failure will allow further production stages during this time to function without any problem. Here, too, the important role of quality is very much visible. If the products come off the machine on the principle "maybe it will be good, maybe not" then we are not able to actually predict what production volume we will need to process the order.
Well, it's impossible to completely get rid of bottlenecks. I think that in an organization where everything goes like clockwork, that is, all machines operating at 100% ideally hit the market demand, every factor is a bottleneck in it. There is always something that limits the system, otherwise it would strive for infinity. And quoting an eminent scientist, only the universe and human stupidity can be infinite. Such a system would have to be highly reliable, however, because an error or failure on any part of the process would cause downtime on other sections, while by maximizing the factor it is difficult to find a buffer later in the event of further unpleasantness.
"Limitations not only inside the company"
As I mentioned earlier in the detection of a bottleneck - you can find them in a place where queues and supplies form. But what if the production process runs smoothly and we are left with excess of finished products? This means that our bottleneck is the absorptiveness of the market and the demand for our products. Here we meet with quite a challenge for the marketing department, which has to climb like an alpinist to the next level and liquify our products. Otherwise, we'll end up with day-to-day stock of excess capacity over demand.
To better visualize the mechanics of bottlenecks, I will present 4 simple schemes where they will occupy different places in the production process. Let's assume that X is a bottleneck, while Y- is not. Both machines are available 24 hours a day, 7 days a week, which gives us a total of 168 hours. The Y machine is able to work faster than X and it takes 120 hours to complete the same part of the service as the X machine.
1) Y -> X
In this case, the machine Y in 120 hours can handle all the needs for machine X. If it does not support other orders that do not pass through the bottleneck, we should let it "rest".
2) X-> Y
At the beginning, probably the most desirable model in which it is clearly visible that this bottleneck is a drummer for the whole process. As much as it comes down from X, that's how many subsequent episodes of the production process will be.
In this case, neither the X machine directly restricts the Y machine, and the Y machine does not overload the machine X. What happens, however, when both machines will produce at their full capacity? Semi-finished products from the Y machine will be more, which will leave them in the assembly hall.
4) Y-> Product A
X-> Product B
The perfect situation, we can produce Y as long as the factory can! Nothing limits us. Are you sure? Once we have established that our bottleneck may be an external factor, and by definition Y has extra production capacity, it may be worth looking at it more closely. Well, each product may have a different bottleneck, it can be melted by other factors. In this case, it may be the previously mentioned demand.
If we produce as much as we can, just to increase our performance indicators, then woe to us. The purpose of the enterprise is in itself to bring profit, not to lift various indicators. The diagrams presented above are very simple. I did not enter here into more complex models to keep a clear and clear message. However, based on these simple variants, you can build any, even the most complex scheme.
Enterprises often face important decisions about which products to set a high priority and whose production should be ended. I hope this task will shed some light on what we were talking about in two articles. First, I encourage you to familiarize yourself with the assumptions of the task, and only then to go to solve it. If you want to receive a new one, unresolved, to check your skills in TA - please contact me.
New technologies are boldly entering the business world. Even smaller entrepreneurs try to optimize their processes and increase their competitive advantage even more willingly. Here are a few examples of "benefits" that a production company can achieve by implementing solutions based on Big Data.
-> the ability to predict the demand for a given good,
-> possibility of locating a bottleneck and earlier obtaining additional powers,
-> fast, real-time decisions about what to produce,
-> recommendations and tips in the event of unexpected problems and downtime,
-> precise description when a given batch will go down from the assembly line before production starts,
-> monitoring the progress and work stage of a given order,
-> the ability to automatically inform the contractor, when production will be at a certain stage,
-> calculation of revenues, costs and profit from a given batch based on processing accounting;
There are many more benefits. It all depends on the partner's requirements and the specifications of his business, because, as well as every one of them, he has his own rules. If you would like to know what benefits new technologies can bring in your company, I invite you to contact me. We are happy to conduct an interview thanks to which we will be able to assess what your company needs and what solutions to implement to improve its performance!