The Deming cycle, also known as PDSA, the “Deming Wheel,” and “Shewhart Cycle”, is rarely used in small manufacturing companies. There are many myths surrounding them, which explains a great fear of implementation, although they are often used unconsciously in various workshops. Many years of consulting experience and working with the ISO 9001 system show that the Deming cycle is often misunderstood or misinterpreted. Read on to find out how to use it in practice to improve production in your manufacturing company.
What is the Deming cycle?
The Deming cycle is a continuous quality improvement model consisting of a logical sequence of four repetitive steps for continuous improvement and learning: Plan, Do, Check (Study) and Act. The method was developed by William Edwards Deming in the 1950s, who stated that business processes should be analyzed and measured to identify sources of variations that cause products to deviate from customer requirements. In addition, he recommended that business processes be placed in a continuous feedback loop so that managers can identify and change the parts of the process that need improvements.
The Four Phases of the Deming cycle
First, identify and understand possible problems. Perhaps the standard of a finished product isn’t high enough. Explore the information available in total. Generate and screen ideas, and develop a robust implementation plan as well as make it measurable.
Once you’ve identified a potential solution, test it safely. This will show whether your proposed changes achieve the desired outcome – with minimal disruption to the rest of your operation if they don’t. Then, as you run the project, gather data to show whether the change has worked or not. You’ll use this in the next stage.
Next, analyze your results against the expectations that you defined in Step 1 to assess changes. If it wasn’t successful, return to Step 1. If it was, advance to Step 4.
This is where you implement your solution. But remember that PDCA/PDSA is a loop, not a process with a beginning and end. Your improved process or product becomes the new baseline, but you continue to look for ways to make it even better.
Simplified version of the Deming cycle
It is possible to simplify the Deming cycle and focus on correction and prevention activities as the essential elements. The version of the ISO norm dated 2015 changes the definition of prevention activities to activities minimizing risk. It means that it is possible to identify the problem and establish its cause when something terrible happens. The next step requires eliminating the grounds, followed by checking progress in the allocated time frame. When the same problem appears again, the whole process is repeated. In this way, we move to a continuous quality improvement model, which concerns not only single actions but global processes improving the functioning of the whole manufacturing company.
How manufacturing software can affect implementation of the Deming cycle?
There are different examples of using manufacturing software in continuous quality improvement models. Read on to find out more.
1. Common mistakes at the shop floor level
Sometimes the same mistake appears regularly at the production hall ex. The machine operator constantly forgets to drill a hole, or the final product has rough edges. The complaints follow because the client has to wait. Is there a solution available? The best way to fix this problem is to identify the cause: whether an operator just forgot about the operation or the task was not typical for the product. Hence it’s missing. How can we minimize this mistake and prevent it from happening in the future? In Prodio, it is straightforward – it is enough to add a photograph to the product. Once the photo is attached, the problematic element can be highlighted even in such a simple program as ex. Paint – draw a line in a bright colour to indicate what should be done, together with the comment: “polish the edges”. Next time an operator starts work on the order automatically, the photo of the product will load. The remark regarding polishing or drilling will be visible, so there is much more chance it won’t be omitted nor forgotten.
There is a similar situation regarding printed production orders. Again, it is possible to write a comment in the product technology, and when an operator picks up a printed barcoded order, they will see that comment straight away.
The next stage involves checking if there aren’t the same complaints, e.g. rough edges or missing holes. But, again, there is no need to make additional changes when everything is alright, and the procedure can be closed.
Other examples of simple corrections and improvements at the shop floor level involve building a database of products accompanied by comments on their production technology. When we focus on adding details once filling technological cards such as machine parameters, temperatures, working time, it is possible to limit the probability of incorrect information added randomly by a worker. They won’t be tempted to “invent” things they don’t know because all required information will be already in the system. When managing production, it is essential to allocate resources and track the status of the project. Ensuring that work is being done consistently and that steps are not being skipped is necessary.
2. Working norms and the Deming cycle
Production scheduling software and the Deming cycle can be used on the micro-scale of particular complaints and globally to control production norms, working time, and production technology.
It is highly recommended to have both precise production technology and technological cards prepared for each product because it will help to monitor norms and establish good quality and standards.
In addition, a good solution is mapping because many different production planning models are based on process thinking. That’s why it is essential to identify processes happening in the company and on its border – between the suppliers and the clients.
Each organization has workflows unique to the type of data being collected and the type of product being delivered. However, these workflows can be generalized into a primary production workflow that consists of steps to create a database and capture or load an initial set of data, perform edits to the data, ensure the data is valid and accurate, and produce digital or hard-copy output. Mapping is designed to streamline each of these steps while remaining flexible to adapt to your business rules and workflows.
The main goal of mapping is data visualization. It is essential when producing a hard-copy product or serving data over the web. Mapping provides the views and visual specification tools for consistent, repeatable, rule-based symbology. You define what symbol or representation should be applied to features based on their attribute combination. It also provides several custom elements, such as graphic tablets, charts, infographics, that allow you to see the whole workflow and processes.
Mapping process of each operation, ex. Lathing, drilling or cutting enables to add suitable time for those operations and production technology. Once the order is ready, we can schedule it within the correct time frame and budget because the production planning software will calculate it. For example, when there are 100 pieces to be made, and each operation takes 30 minutes, the program will count the total time needed to complete the order, which will help create a production master schedule.
The order is scheduled, and workers at the production hall can now clock in with barcoded cards or RFID key-fobs to start work. If they use production planning software such as Prodio, each operation is registered in the system, so it is possible to track real-time progress. It is also clear how much time was needed for a particular operation.
Mapping allows us to identify all processes and operations where we waste time and money because it is easy to identify potentially dangerous places where mistakes can happen. Once there is a clear view of the situation, the weak links can be eliminated and order restored.
By analyzing working time data, we can improve productivity. When it’s clear that the time was exceeded (at production), it is possible to use checking – the third step of the Deming cycle. For example, let’s say that we planned 50 working hours to complete the order, but in the end, there were 65 hours spent, which gives us 15 extra hours. It is a significant amount of time and a great discrepancy between planned and real working time. There can be a few different corrective measures at this stage, but it is evident that something has to be changed ex — elements in the production technology. The client pays only for 50 hours, not for 65, so either the norms have to be increased or the pricing for the following order.
Once the order repeats, it is possible to check if the corrections proved successful and production was optimized. It is good to review progress regularly and make adjustments accordingly. Implement what’s working, continually refine what isn’t, and carry on the continuous improvement cycle. The Deming cycle is a constant loop of planning, doing, checking (or studying), and acting. It provides a simple and effective approach for solving problems and managing change. The model is helpful for testing improvement measures on a small scale before updating procedures and working practices.
The approach allows a planning phase in which problems are identified and understood, and a theory for improvement is defined. Potential solutions are tested on a small scale in the Do phase, and the outcome is then studied and Checked. Sometimes it is necessary to go through the Do and Check stages as many times as necessary before the complete, polished solution is implemented in the Act phase of the cycle.
Another excellent example of global use of the PDCA cycle can be measured by introducing primary production KPIs. There aren’t many manufacturing companies that monitor the efficiency and profitability of production processes. The working time spent can be a good example. When the time spent at work is registered in the RWT system (registration of working time), it is possible to say how much time was spent effectively (operation on different machines), not just lazing around. Once we decide that unproductive time should be minimized to 10 per cent of the total time spent at work, we can start monitoring it and analysing the results. When production planning software such as Prodio indicates that 20% of the time was spent unproductive, it is possible to start correcting measures. We can identify the cause of the problem and quickly fix it. Analogically similar solutions can be introduced regarding machine breakdown, change over time, bottlenecks or delays thanks to a significant number of different production KPIs. In the following article, you will learn more about using various data from the Prodio system and analyze it.