Process simulation for realistic optimisation

A customer-specific, analogue learning factory followed by digital, dynamic simulation translates into a method that a company has developed to enable comprehensive and realistic optimisation of production processes.

What if …? This is the question every entrepreneur and business asks almost on a daily basis. How would a change in production processes affect my service level? How will my processes react to a 10% rise or fall in workload in the second quarter? Are orders allocated efficiently to the machines? Does it make sense to invest in a new plant? And if so, what minimum/maximum capacity do I need?

Testing during production rarely an option
Usually there is no lack of ideas on how to optimise the processes – what is lacking, more often than not, is the opportunity to test out suggested concepts and procedures. There is no easy way to make changes to production during ongoing operations in order to evaluate proposed improvements, nor is it a good idea to blindly buy new machinery. Pilot runs of potential optimisation measures, therefore, are not always wise or possible for reasons of time, cost or risk factors.

Generally, this is explained by the fact that the human brain easily understands linear cause-effect relationships. This is instilled in us as early as throughout our schooling – and often shapes our basic understanding of correlations – in that assumptions are made at every juncture under the otherwise same conditions, whereby causal relationships are organised into an easy-to-digest logic. If the price rises, the demand will fall! If process 1 is faster, the service level will improve!

Real systems often do not respond as expected
Real systems, however, frequently present us with complex reactions that typically are not immediately obvious and that – in the face of all apparent logic – change the expected outcome entirely. A simple example of this, which is familiar to us from real life, is that abolishing the speed limit on the motorway can potentially increase the risk of tailbacks at a completely different point, even on some ordinary road nearby. Translating this question into production, we need to ask: “Why should I increase my average throughput time if I accelerate certain process steps?”

The reason for this is that processes today are a complex system of subprocesses, where changes in feedback loops can easily, and often, it seems, unpredictably, be transferred from one element of the system to another, even separate, element up- or downstream. Problems in the system usually arise where typical optimisation measures are put in place as a quick fix, to put out fires as and when they occur, without systemically analysing the actual causes and effects.

A further problem is that companies in many cases lack the necessary foresight for improvement measures. Driven by operations, employees and often the management are mostly focused on their own areas, making it difficult to embed a readiness to change in this kind of environment. How could you? Processes and problems outside people’s own responsibility are rarely taken notice of or questioned, due to a lack of time, but also pure indifference. However, this is exactly the point where a systemic mind-set must come in, and with it, the basis for enabling effective and efficient improvements, ideas and solutions. Employees must be “opened up”, sensitised and mobilised, they must recognise – ideally, even experience for themselves – how problems arise with cross-departmental or systemic effect. Only then can readiness for change develop! Unfortunately, small to medium-sized businesses in particular rarely have appropriate concepts in place or the means to bring about this readiness for change.

For these cases, Kerkhoff Consulting GmbH has developed the Performance Centre as a learning factory where companies can simulate specific problems in a “protected environment”, identify necessary changes and on this basis develop appropriate measures. The focus is deliberately on practical experience, in other words, rather than presenting participants with general theories, they themselves actively simulate relevant situations. Of particular importance here is that they learn to take the perspective of others – for example, the purchasing agent assuming the role of staff in production planning, assembly or quality management, and vice versa, thereby enabling them to think outside the box.

Simulating specific problems in the learning factory
The learning factory devised by Kerkhoff differs from a typical learning factory and competing concepts particularly in that every detail is geared to the specific context of the client. It is customised by process experts according to the needs of each user. Every unique aspect is taken into consideration, with processes simplified to the point that a transfer into reality is easy. And since the user gets to keep the “factory” at the end of the project, they can always continue as necessary to tweak their own processes.
The use of a learning factory can be a critical success factor in the project. Not only does the learning factory provide an overview of the corporate processes, it also helps to identify relevant need for change. In simulations, staff are encouraged to remedy process deficits themselves, thereby increasing motivation as a result of high success rates of the work carried out in the learning process. Another advantage of the Performance Centre is that changes are repeated until the client is satisfied. This eliminates scepticism towards changes, as opportunities and risks are identified before they are implemented in the company. As a result, the learning factory provides a platform for effective and sustainable changes.

Enhancement of productivity, output, quality and planning accuracy
Based on a holistic approach, these optimisations produce a number of positive effects for the whole company: conflicting goals can be reconciled, resulting in enhanced productivity, output, quality and planning accuracy. At the same time, throughput times and setup costs are reduced and production costs, logistics, inventory and working capital minimised.

In reality, though, processes are often a great deal more complex than the setting and the simplified representation of processes simulated in the learning factory. Aware of this, Kerkhoff experts can therefore digitally reproduce the customer-specific process in an IT system. The focus of this dynamic simulation is on a specific issue (such as reduction of the throughput time, with less rejects), thereby achieving a balance between realistic accuracy and sufficient abstractness of the model while stochastically accounting for diverse influencing and interfering factors. The ideas and options worked out can subsequently be tested for effect based on real company and market figures, and statistically supported.

This ensures that only those optimisation measures and changes are piloted and implemented that promise the highest success rate for the user. There is a further benefit in that sensitivity analyses allow users to test the reaction of the target process to internally or externally initiated changes.

Many small to medium-sized companies today deal with complex processes in their production. A video at this link demonstrates the problem and potential solutions.

03 March 2017 | Authors / editors: Wilhelm Woldenga / Reinhold Schäfer