A guide to making useful Factory Simulations

A guide to making useful simulations. The picture shows a treasure map with a big red cross.

Making useful factory simulations

At Production Support 56 our mission is to cut out the waste and help businesses add value to their production processes by creating useful factory simulations. This can be of an existing factory or a new factory design. This article describes the what, the why and most importantly the how we help you make a useful simulation.

 

There are two brief initial sections on what and why, but if you already know this then just skip to the how.

What is Factory Simulation?

A factory simulation is a model of your production facility made in discrete event simulation (DES) software. With the model and software, you can run simulations of your factory operations using different scenarios. For our simulations we use Simul8 software.

 

The model is constructed by describing the individual elements of the factory, how these are linked, resource availability and decision making. So, for a piece of equipment, you’ll define parameters that are important to your simulations such as, how many you have, what is the capacity, cycle time, start-up period, change over times, resource requirement, material routing, decision making etc. Doing this you can build complex models using basic building blocks. The amount of information required to build a useful factory simulation depends on the question you want answering, and that is where we help. We make sure only the relevant data is captured and to a sufficient accuracy. It is often less demanding than you expect.

 

Below shows a simple example of a factory simulation.

 

In this article I use the words model and simulation interchangeably. However, when referring to the model I normally mean the logic that describes the factory, and when referring to the simulation I mean a specific scenario being run on the model. Other simulators may use a variation on this definition.

 

So why would you go to the effort of building a simulation? It takes time and energy to create a simulation of your production processes, so why bother? Well, there are whole host of reasons and applications for the use of simulation in manufacturing. In this article we focus on two applications: designing a new plant and improving an existing plant.

Why is simulation useful for factory design?

When you are designing a new factory or production line, creating a simulation as part of the project can really support the design process. Here are some of the key benefits of using simulation to support designing new plant:

  • The project is already collecting most of the information required to create a useful factory simulation. It is a small step to use this to create a model and test your design at every stage. It is easy to over think a design and miss ‘obvious’ issues. The model provides a sense check to the design and can prevent oversights becoming built-in.
  • The model provides accurate dynamic information about your design. This allows you to continuously test drive and validate your design against the design specification.
  • It aids the production of a technical scale-up plan to support development of the design. Creating a model highlights the key information requirement for the detailed design. This can be used to plan equipment trials, pilot production and supplier conversations.
  • Reducing capital costs. Using the model during the design phase, you can identify under-utilised and over-sized equipment, allowing you to confidently reduce the capital costs.
  • Identify improvement opportunities. A simulation of a manufacturing facility design, allows your team to identify bottlenecks, wasted resources and high stock levels. This can then be resolved at the design stage, and not after the facility has been built.
  • Cost-benefit analysis. The benefits of optional extras can be quantified, and real value can be identified early, instead of ad-hoc changes being added later.
  • Determining manufacturing resilience. Using scripted events, you can test the resilience of the production facility against breakdown of key equipment.
  • Communicating the design to the stakeholders. The animated representation of your design makes it easy to communicate your plans to stakeholders. Earning valuable buy-in and useful feedback.

For more details on this read our ‘Better Factory Design’ article or take a look at our Simulation for Factory Design service page.

Why is simulation useful for improving existing factories?

When you manage an existing production facility, a simulation will let you test changes virtually without impacting real production. When you have thoroughly tested and optimised your ideas you can confidently implement them in real life. Here are some key benefits of using simulation to improve existing production facilities:

  • The simulation model can be rapidly re-run with different scenarios as many times as you like. This allows improved forecasting so you can optimise your production plans.
  • You can define and optimise your labour availability and skills matrix against your production plan. Meaning you can not only check you have sufficient staff for your plans but also that you have the right skills in the right place at the right time.
  • Different products often have different resource requirements and simulation allows you to optimise your product mix to maximise throughput and utilise resources efficiently.
  • Some operational issues only occur when certain activities coincide. Simulation helps you create those scenarios and helps identify causes of issues.
  • You might have some downtime of key equipment planned. Simulation software allows you to script these events and test out your contingency plans.
  • You probably have many ideas of how to improve your production facility. Simulation lets you test your improvement ideas, quantifies the improvement, calculates a return on investment, helps optimise your improvement, and identifies the best sequence to implement them.

We will discuss the benefits of simulation for improvements in more detail in a future article.

How to make useful factory simulations

We always start with the end in mind. What are we trying to achieve? What operational issue are we trying to resolve? We work with the client to define the question and gather the important information; we then make the model with support from the client. Finally, we interrogate the simulation with the client, to wring out as much useful information as possible.

 

Below summarises the typical stages of one of our simulation projects. We refer to ourselves as PS56 (Production Support 56). This process has been optimised to us, if you are interested in making models yourself you may want to adapt our process to your skill set and resources.

 

  • The Question- Define the goal and purpose of your model. Understanding the key levers and outputs. (PS56 & Client, 2 hours). Once we have a good idea of the scope of the project we can provide you with a proposal.
  • Data gathering- We create a process map, gather product and resource information. At this stage we need to know how your production processes fit together, not necessarily all the fine detail. (PS56 & Client, half a day).
  • Draft model- We create a draft model of your production processes along with a questionnaire of the key information required. (PS56, off-site)
  • Validation of draft model- We populate the draft model with updated information from the client. We then run the simulation with the client with the intention of validating the model and confirming the key levers and outputs. (PS56 & Client, 2 hours)
  • Detailed Model- Now we are confident that the structure of the model is correct, we add all the required functionality and detail. We use a spreadsheet to configure the simulation scenario. A second spreadsheet is used to manage the process performance data, such as cycle times, this can easily be updated as new information is generated. (PS56, off-site)
  • Model interrogation- This is the big one. The model is run with all the key stakeholders, such as operators, managers, engineers, procurement, sales and finance. Anybody with an interest in making the process better. They get to ask the model lots of ‘what-if’ questions. This should answer The Question and more. (PS56 & Client, half a day).
  • Report out- All this effort will be wasted unless it delivers something. The key findings from the model are reported plus any other insights found from the model interrogation. The client is supplied with a copy of the model, video clips and a summary of the key findings.

In the rest of the article, I will describe these stages in more detail, and describe the experience and expertise we bring to a simulation project.

The Question

This is where all good simulation projects should start. Having a clear operational question in the beginning gives purpose to the project and helps focus efforts. The Question will determine what information needs collecting. From The Question we work out what information is needed in the start to complete the draft model. The rest of the information required to finish the model can be collected on a longer time scale.

 

The Question will help define the key output in terms of a metric. Such as, throughput in kg/hr or lead time in terms of number of days. It will also help identify the key levers- the variables that you have some control over and affects the key output. Such as, capacity of a storage silo in tonnes or the changeover time of a press in minutes.

 

Some examples of questions are:

  • Will this design meet the target throughput?
  • Where can I save on capital costs without impacting capacity?
  • What are the required staffing levels and operational costs?
  • Which improvement project has the biggest impact?
  • What is the most efficient product mix?
  • How long will it take to make this order?

Of course you will have secondary questions, nice to haves and future developments. These are worth considering as some might take very little extra effort and produce a more useful simulation.

We’ll use these requirements to produce a practical scope to meet your needs, and a proposal detailing the time scales and costs.

For more discussions on secondary questions check out our ‘Factory simulations: It’s all in the preparation’ article.

Data gathering

This is where PS56’s process improvement skills make a big difference, as it is a very similar approach to an improvement project. We start by drawing a process map with the process owner- getting the post-it notes out and doing it properly! This gives clarity to all involved about what the process looks like, and what will be modelled. It also provides you with the structure of the model and is useful for identifying what information is readily available and what needs to be gathered.

Even if there are process maps and process flow diagrams available it is still important to go through this exercise. We often find that existing documents are out of date and don’t tell the full story. When we go through this exercise it is important that we have at least one person who knows the process well.

At this point it is important to get the structure correct- the order in which activities are linked- but not necessary the performance characteristics of the individual activities. The reason being is that it can be tricky and messy to change the structure of a model, but it is easy to change performance data such as cycle times.

Understanding the key outputs and levers of the simulation helps you gather the right data. For example, if staffing is not an issue, we can assume labour is always 100% available and we do not have to worry about understanding the shift pattern and skill matrix.

Draft model

We can now get started with the simulation. The job here is not to create the finished article but create a draft that the client can review. The purpose of the draft model is to enable us to validate the material flow.

So, normally we would create the model’s structure, get all the activities in the right order and with the right routing, make sure all materials and information are flowing correctly. Adding in to the model the key output and levers.

The model does not have to be pretty, but it does need to be legible. It does not have to be user friendly, but you need to be able to drive it. Only the features required for validation need to be included. Once this structure is confirmed it is easy to add in all the details and data management features.

Whilst we are constructing the draft model, we make a questionnaire of all the missing data and a list of features to be included in the finished model.

Validation of the draft model

The initial aim here is to check with the client that the model describes the material flow correctly. To do this we demonstrate the model a few times with the client, paying attention to how the material flows through the production process. Do they recognise it? Does material move and hold-up as they expect?

If there are any errors, if possible, we correct them there and then. If not, we make notes and make the corrections after the meeting. This is where the age of on-line meetings has made the job so much easier. We can just book a 30-minute meeting when the model is updated and run through the changes. We don’t have to decamp to somebody else’s meeting room for a 30-minute chat.

Once everybody is happy with the draft model, we can then run through the questionnaire of missing data and talk the client through the list of planned features. At this stage the client will likely have some changes and additional requests for us. If they are easy to incorporate- then we add them, if they are not relevant- we leave them out for another project, if they are important but time consuming, we update the proposal or swap it for an unwanted feature.

Detailed model

Now it’s time to put it all together and create a useful factory simulation. I’m not going to dwell here as this is a full topic in its own right. In a future article I will discuss how we structure a model so that it can be constructed by a team, and easily updated and developed.

With all the groundwork in place this is just about grinding through the modelling. There will be some communication to clarify aspects, but nothing major. Where accurate data is not yet available, we will use variables with best estimates that can easily be updated as better information is acquired.

Model interrogation

From the project point of view, this is the important bit (and the most interesting). The preparation and simulation building were just to get us to this point. The model should be able to answer The Question and more. We often run through the finished model with the process owner first. Making sure it still looks right and gives sensible answers. It is also an opportunity to identify what other questions are going to be asked.

Now its show time, the big meeting with the client. We invite the key stakeholders, allow at least two hours, and make sure there is plenty of tea, coffee and biscuits. We slowly step through the model, following the material flow, showing them the key levers and how this affects outputs. We’ll run through the important features and any assumptions that have been made.

Once the stakeholders understand how the model works, we get them to ask ‘what if’ questions. This will involve running the simulation with different scenarios and assessing the output. The questions often push the model’s flexibility to the limits, and sometimes we may have to alter the model on the fly. The stakeholders normally run out of questions before the model runs out of answers. During the meeting we make sure we record any interesting findings.

Reporting out

The hard work is done. We have created a useful simulation, the stakeholders have interrogated it, and gathered insight and answered their operational questions.

We provide the client with a copy of the model, video clips and a summary report of the interesting findings.

It does not have to end here. The client can access the model longer by using a viewer or buying a software licence. With this they can continuously access and develop the model to provide them with ongoing support to their decision making.

Why use Production Support 56 to create your Simulation?

Production support 56 is made up of James Deane and me (Michael Stephenson), we have both worked in the manufacturing and technology industries for over 20 years. We are both process people and think in process. We have both held different job titles, but it has always been about the process. Our careers have covered process improvement, process development, process control, process safety, process design, process scale-up and process modelling. There seems to be a pattern here.

So why do we use simulation? That’s easy, process mapping and process modelling are vital tools when working on processes, and simulation is a dynamic process map-model. It is the perfect tool when you are working on a series of processes and want to understand how they work together. That is the reason why we have been creating our own simulation models for over ten years. It is our ability to think in process that allows us to accurately describe processes and transcribe them to the simulation, resulting in a simulation that accurately reflects real life process behaviours.

We have both worked on many engineering design projects as the process expert and now as the simulation expert. Our experience and expertise, allows us to create really useful simulations, which add lots of value to the design process. Our day job is working with Operations, to make improvements, find better processes or increase throughput. Simulation allows us to virtually test and communicate changes before they are implemented, significantly improving success rates.

It does not stop with the making of a simulation, you then have to do something with it. Our experience in manufacturing allows us to ask pertinent questions of the simulation, and our process improvement skills allows us to thoroughly analyse the generated data. The final step is to sell the ideas. The simulation makes an excellent communication tool backed up by our professional presentation and reporting skills.

 

How do we get started with Simulation?

If you are new to Factory Simulation and can see how it would be helpful to your business, the easiest way to start is to get in touch and we can help you go through the process. If simulation is a skill you’d like to bring in-house, then we can work closely with your team and train them as we work through that first project.

 

For  more information, call me, Dr Mike Stephenson, on 07305 070155 or contact me via email [email protected]